Skip to main content

Full text of "Conservation Of Antiquities And Works Of Art"

See other formats




I Ot.ass 

j Call No. 

i- - — - 

D a.A 79. 


Oxford University Press, Amen House, London E.C.4 



Treatment, Repair, and Restoration 







My previous books. The Preservation of Antiquities and 
The Conservation of Prints, Drawings, and Manuscripts 
(now out of print), were written at the request of the 
Museums Association, and it gives me great pleasure 
to be associated with them once again in the present 
work, which has been accepted as their official text- 
book on the conservation of museum objects. 

H. j. p. 



Call Ko 





This book is concerned with the broad field relating to the restora- 
tion of antiquities and works of art, and with their subsequent 
conservation. It is intended as a handbook for the collector, the 
archaeologist, and the museum curator, and as a workshop guide for 
the technician. 

As collectors know only too well, the acquisition of objects is 
but the first step towards their incorporation in the collection. In 
order to be able to appreciate and study the objects, it is usually 
necessary to clean, restore, and repair them, and always necessary to 
maintain a suitable environment which will ensure their stabihty 
whether in storage or on exhibition. In the following chapters simple 
instructions are given for cleaning and preservation, and the collector 
with a practical mm of mind who desires to carry out for himself the 
methods described can do so without any special technical training. 

The archaeologist will find interest in the methods that science has 
to offer for the restoration of antiquities firesh firom excavation and 
for revealing evidence of value to him in his researches. Many 
examples are given throughout the work, showing how imsuspected 
facts are brought to hght during the normal course of laboratory 
investigation and treatment. 

The requirements of the museum curator, who is not always in a 
position to call in the museum scientist, have been particularly con- 
sidered, and it is hoped that this pubUcation, which deals with the 
numerous causes of deterioration in museum objects and their treat- 
ment, wiU enable him to detect and arrest decay in its early stages, 
and also to carry out the simple cleaning operations that so often add 
interest and value to the material in a collection. 

The subject-matter is necessarily very varied, based as it is upon 
the day-to-day problems that are presented in the Research Labora- 
tory of the British Museum — problems relating to books, prints. 



drawings, manuscripts, textiles, coins, objets d' art, ethnographical speci- 
mens, and antiquities of all kinds. From this wide range of material 
examples have been selected that illustrate common types of deteriora- 
tion, and an attempt has been made to recommend from among the 
various methods of treatment available those that have proved most 
effective and are at the same time relatively easy to apply. In order to 
preserve the balance, however, detailed descriptions are included of 
some of the major tasks of restoration that have been carried out. 
This has been done partly for the sake of interest, and partly to 
emphasize that it is impossible to prescribe for all contingencies; each 
specimen that is submitted for treatment presents its own individual 
problems, and standard methods of treatment may have to be adapted 
or new methods devised before a satisfactory restoration can be 
achieved. Whichever method of treatment is chosen, it should be 
apphed so as to yield results that lie between the extremes of over- 
and under-cleaning, the aim being to realize the golden mean which 
wiU satisfy at the same time the requirements of science, art, and 

The special problems of the picture gallery have not been over- 
looked. In common with the museum, the picture gallery is vitally 
concerned with the stabihty of materials and methods of conserva- 
tion, but the restoration of easel paintings is a highly specialized 
undertaking, and while instructions are given for carrying out some 
of the simpler studio processes, it is not the intention of the author to 
encourage the amateur to attempt intricate operations on valuable 
material. Such work is for the professional artist technician — one 
who has practical experience based upon a knowledge of the methods 
used by the Old Masters in the different schools of painting. It is 
important nevertheless that the collector or curator of paintings 
should himself be familiar with all aspects of picture conservation. 
He wiU then be able to discuss his problems in a knowledgeable way 
with the restorer, and take a personal interest in any treatment that 
may be required. 

All the processes described herein have been tested, most of them 
at first hand, by the author. Many are standard methods that have been 



handed down through several generations, but some are offered for 
the first time. While the methods recommended are all based upon 
scientific investigation, the book is not written for the scientist. On 
the contrary, a conscious eSbrt has been made to write for the non- 
scientist who has the responsibdity of caring for art treasures. By 
presenting the material in this way it is hoped that the work will be 
of service to a wide range of readers not only in the museum world, 
but also in the home, where, indeed, many of our greatest treasures 
are still to be found. 

I am greatly indebted to the members of my staff who have put 
their specialized knowledge at my service. In particular, my thanks 
are due to Dr. A. E. Werner for reading the manuscript and making 
many valuable suggestions; to Mr. R. M. Organ (metals); to Miss 
Mavis Bimson (stone, ceramics, and glass) ; to Miss Sylvia Schweppe 
for her help in collecting and collating the material; and to Mr. L. H. 
Bell for his help with photography. 

H.J. P. 





introduction: the influence of ENVIRONMENT I 


















































IV. Pearson’s square for simplifying dilution 

















Decorated Ivory Plaque from Nimrud (pth cent. b.C.) frontispiece 

After Treatment 

1. Portrait Vase from Peru (Mochica Culture) facing page 20 

2. Section through a Piece of Ox-Hide 21 

3. Leather Bindings and their Treatment with Leather Dressing 21 

4. Decay of Leather Bindings by Chemical Action 28 

5. Reaction of Vegetable-Tanned Leathers to an Accelerated Ageing 

Test 28 

6. Egyptian Mathematical Leather Roll (c. 1700 b.c.) 29 

7. Egyptian Papyri, before and after UmroUing 48 

8. Method of Mounting Birch-Bark Manuscripts 49 

9. Manuscripts recovered from a Buddhist Figure 64 

10. Drawings by Leonardo da Vinci 65 

11. Stages in the Removal of a Stain from a Printed Book 80 

12. Water-colour Drawing by Guardi 81 

13. Fragment of Chasuble from Lincoln Cathedral (13th cent.) 96 

14. Maniple of St. Cuthbert from Durham Cathedral (loth cent.) 97 

15. Pyx Cloth from Hessett Church, Suffolk (15th or early 16th cent.) 112 

16. Apparatus for Fumigating Objects with Hydrogen Cyanide 113 

17. Apparatus for Impregnating Objects with Molten Wax 128 

1 8 . Specimens of Waterlogged Wood, Treated and Untreated, after 

Drying 129 

19. Sword and Scabbard from Stanwick (Iron Age) 144 

20. Decorated Ivory Plaque from Nimrud (9th cent, b.c.) Before Treat- 

ment 145 

21. Diagram of the Cross-Section of a Panel Painting 160 

22. Silver Lamp from Ur of the Chaldees (c. 3000 b.c.) 161 

23. Gilt-Bronze Ornament from Sutton Hoo Shield (7th cent.) 176 



24. Gold Head of a Bull from Ur of the Chaldees (c. 3000 b.c.) 

25. Section through a corroded Silver Bead 

26. Section through a corroded Bronze 

27. Silver Cup from Enkomi, Cyprus (r. 1400 B.c.) 

28. Silver-Gilt Buckle (Frankish, 7th cent.) 

29. Silver-Mounted Glass Beaker (ist cent, a.d.) 

30. Hoard of English Silver Pennies (Edward I and Edward II) 

31. The Emesa Helmet from Damascus (ist cent, a.d.) 

32. Hair-Hke Cracks on Speculum Mirror (Han Period) 

33. Oyster-shell Pitting on Speculum Mirror (Han Period) 

34. Egyptian Bronze Group of Isis and Homs (Ptolemaic) 

35. Bronze Bust of Ethiopian Queen (yth-tith cent, b.c.) 

36. Bronze Shield-Boss from Anglesey (Early Iron Age) 

37. Fragmentary Bronze Bowl mounted on Perspex Model 

38. Fragmentary Bronze Bowl mounted on Perspex Rings 

39. British Museum Research Laboratory: Part of Metal Conservation 


40. Stages in the Disintegration of Leaden Objects 

41. Pewter Medal of the Capture of the Bastille 

42. Iron Steelyard with Leaden Weights (Romano-British, c. 3rd cent. 


43. Sword-Hilt from Sutton Hoo (7th cent.) 

44. Shield-Boss from Sutton Hoo (7th cent.) 

45. Helmet from Sutton Hoo (7th cent.) 

46. Fragment of Iron Sword with Inlaid Ornament (Roman, c. 200 a.d.) 

47. Iron T’suba iulaid with Silver Qapanese) 

48. Iron Buckle inlaid with Silver (Frankish, 7th cent.) 

49. Corroded Iron Purse-Mount: Decoration revealed by X-Radiography 

(Merovingian Period) 

50. Limestone: Egyptian Sculptor’s Trial Piece (Ptolemaic) 

51. Carved Stone Figure from Mexico (probably Huaxtec Culture) 

52. Marble Head of Mithras, fromWalbrook, London. 

53. Marble Sculpture from the West Frieze of the Parthenon 






























54. Cuneiform Tablets of Clay 352 

55. Apparatus for Mounting Prints 353 


1. Hygrometric Chart 6 

2. Synchronous Variations in the Temperature and Relative Humidity 

of the Atmosphere in Different Locations 7 

3. Causes of Damage to Museum Objects 15 

4. Making Mounts for Prints 90 

5. The Vacuum Desiccator 152 

6. Equipment for Electrolytic Reduction 194 

7. Use of Needle in Mechanical Cleaning 202 

8. Section of Corroded Bronze Spear 237 

9. Air-tight Case kept Dry by SiHca Gel 245 

10. Use of Ion Exchange Resins for the Treatment of Leaden Objects 263 

11. Sand-blast Apparatus for Field Use 322 

APPENDIX XIII. Adapting Mains Electric Supply for Electrolytic 

Reduction 360 


Pi. 10 is reproduced by gracious permission of Her Majesty the Queen. 
Acknowledgements are due to the Illustrated London Neivs for contributing the 
frontispiece; The Times for permission to reproduce Pi. 52B; the Laboratoire Central 
des Musks de Belgique for Pi. 17; thte British Leather Manufacturers’ Research 
Association for Pi. 2; Nickeloid Photography for Pi. 16; B. S. Cron for Pi. 48; 
Walter Hege for Pis. 53 a and b; for permission to reproduce all other illustrations 
the author is indebted to the Trustees of the British Museum. The drawings were 
contributed by Mr. C. O. Waterhouse, M.B.E., of the British Museum. 



The condition of an antiquity or a work of art depends on the 
material of which it is composed and on the environment it has 
experienced during its hfe history. Its ultimate survival will depend, 
not only on the appHcation of any treatment necessary for its pre- 
servation, but on its being provided with an environment in which 
it will remain stable. Environment plays a major role in conserva- 
tion, and for this reason any study of the innumerable aspects of the 
conservation of museum objects must be preceded by a general con- 
sideration of the effects of environmental change — in particular, 
change of temperature and relative humidity. ^ 

The problem relates both to the effect of change of environment 
on objects immediately after excavation, and to the question of creat- 
ing a museum chmate suitable for the miscellaneous collections 
housed in museums and art galleries. 


Before archaeologists were aware of the effect of sudden changes 
of relative humidity, it was not an vmusual experience for them to 
open up a tomb in which the contents were in perfect condition, 
only to see them shrink and warp and sometimes even turn to dust 
when exposed to a change of atmosphere. 

‘ Relative Humidity (R.H.): The relative humidity of a sample of air is the ratio of 
the amount of moisture actually present (w) in any volume, to the amount needed for 
saturation (M) at the same temperature, 

i.e. R.H. = (m/Afx loo) per cent. 

Air is said to be saturated when it is unable to take up any more water in the form ot 
vapour. On cooling saturated air, water is at once deposited in the form of dew, because 
the air is unable to carry the same amount of water at the lower temperature. 

B 6167 



The principle behind this is that when objects are buried for a long 
time under conditions that are reasonably constant, they tend to 
attain a state of equihbrium with their sinroimdings. When excava- 
tion takes place, the materials are submitted to an entirely new 
environment, the equihbrium is upset, and some objects suffer pro- 
foundly. When a tomb is opened, the access of cold air increases 
the relative humidity, even to the extent of causing deposition of 
moisture, while the entrance of hot dry air causes a reduction in the 
relative humidity. Such variations give rise to dimensional changes 
in organic material — due either to the absorption of moisture or to 
the loss of moisture — and such dimensional changes manifest them- 
selves in swelling or in shrinking and warping. 

A similar kind of problem arises when objects that have been 
excavated from damp soil are dried quickly instead of being allowed 
to adapt themselves gradually to the new environment above ground. 
In the case of organic materials, the sudden loss of moisture may 
cause shrinkage and severe distortion, and in the case of stone and 
pottery, crumbling. In order to prevent such deterioration, it is 
essential, on excavation, to keep objects in a cool place out of the 
sun, where they can give up their moisture slowly. 

When objects having a cellular structure, such as wood, bone, and 
ivory, have become waterlogged, the problem, on excavation, is 
accentuated. It is essential to keep such objects immersed in water 
until such time as they can be packed in wet cloths and transported to 
a laboratory where they are treated by one or other standard process, 
or, in some cases, merely dried out slowly under controlled condi- 
tions so that shrinkage and distortion are reduced to a minimum. If 
allowed to dry too quickly, the cellular structure may collapse, and 
in the event of this happening no amount of resoaking in water will 
restore the original shape. An extreme case is that of damp organic 
material having an impervious covering that must be preserved, e.g. 
Chinese lacquer. With objects of this kind, the most careful treat- 
ment is not always a guarantee of success as the following example 
win show. In 1949 four trays made of very thin lacquered wood were 
excavated at Chang Sha in a waterlogged condition. They were 


immediately packed in zinc-Hned boxes containing damp wool and 
sent by air for treatment in the Research Laboratory of the British 
Museum. There, in a series of closed vessels containing air of suc- 
cessively decreasing relative humidity,* they were dried for twelve 
months. Although all four objects received exactly the same treat- 
ment, only three of the trays could be preserved; the fourth con- 
tracted, warped, and finally fell to pieces. The reason for this anoma- 
lous behaviour may have been due to the nature, quahty, or direction 
of cut of the wood used, in making this particular tray; on the 
other hand, the tray may have had a more chequered history than 
the others, in that it may have been subjected to periods of inter- 
mittent drying while stiU. in the ground or even to partial drying 
when it was excavated. 

Objects buried in salty ground are exposed not only to moisture 
but also to the action of salts dissolved in the ground water. When 
damp materials containing soluble salts are excavated, the water 
evaporates and the salt crystallizes, and in so doing weakens the object 
and often disrupts the surface (Pi. i). In this case, provided a 
source of salt-free water is available, it is generally advisable to wash 
out the salts before crystaUization takes place. 

Metalhc objects also become contaminated in salty ground. Al- 
though they corrode in the ground there may come a time when 
they attain a state of equiUbrium with their surroundings and the 
process of corrosion is brought, virtually, to a standstill. On being 
excavated and exposed to a new environment, corrosion may break 
out afresh as they proceed to adapt themselves to the new conditions. 


While antiquities may suffer deterioration through exposure to 
the sudden change in temperature and humidity consequent upon 
excavation, certain works of art may suffer as severely by exposure 

‘ At a temperature of 70° F. the R.H. of the air within a closed space containing 
saturated aqueous solutions of the following salts b as follows; potassium nitrate 94 
cent., potassium chloride 86 per cent., sodium chloride 76 per cent., sodium nitrite 
66 per cent., magnesium nitrate 53 per cent. 



to ordinary everyday atmospheric changes. Easel paintings, for 
example, whether on panel or canvas, are very sensitive to changes in 
the atmosphere, and in particular to varying relative humidity. When 
a painted panel is subjected to variations in relative humidity, the 
wood undergoes sympathetic dimensional changes which impose 
rhythmical strains on the painting ground (gesso) and the paint layer. 
As a result of this movement, cracks and bhsters may develop, and 
the paint may flake off. Before the Second World War a technician 
was employed for eight months in the year to deal with such defects 
in pictures at the National Gallery where the atmosphere was not 
conditioned. On the other hand, in the repository which housed the 
same pictures during the war and where the air was conditioned to 
58 per cent. R.H. at 63° F. the work of the technician was reduced to 
approximately one month during the first year, progressively less in 
the following years, and finally after five years his visits became a 
mere formaUty as there was no longer anything for him to do. The 
significance of the relative humidity factor in the conservation of 
pictures was further emphasized by the fact that, when the pictures 
were returned to London, they behaved in the same way as before 
the war, the paint flaking off as badly as ever in rooms that were not 
air-conditioned. I 

Similarly, in a large repository used during the war to house the 
British Museum and Victoria and Albert Museum collections (anti- 
quities, books, textiles, prints, drawings, manuscripts, furniture, &c.), 
the air was controlled at the constant figures 60 per cent. R.H. at 60° F., 
and not a single case of deterioration was recorded in this repository. 

No more striking evidence could be brought forward to illustrate 
the paramount importance of maintaining temperature and atmo- 
spheric relative humidity constant at an appropriate level in museums 
and picture galleries. Where there exists an air-conditioning plant 
incorporating a dust extractor, as is customary in modem installations, 
conditions are ideal. Under such circumstances there is no longer any 
risk of damage by atmospheric pollution; soot particles are absent 

' Keeley, T. R., and Rawlins, F. I. G., ‘Air Conditioning at the National Gallery, 
London’, Museum, 1951, 7, p. 194. 


and it is safe to remove glasses from frames so that pictures may be 
enjoyed without the comphcation of specular reflections. 

While for the majority of collections such strict atmospheric con- 
trol as 60 per cent. R.H. at 60° F. is not essential, there is everything 
to be gained by fixing an upper and lower limit of relative humidity 
and temperature and ensuring as far as possible that conditions are 
maintained within these predetermined limits. This involves keeping 
records (see Appendix VII) and having the means of controlling 
heating and ventilation. 

A graphic representation of the relationship between temperature 
and relative humidity is showm in Fig. i, from which it will be seen 
that if at 70° F. the reading is 50 per cent. R.H. (curved hne) — con- 
ditions that might well be recorded by day — it only requires a fall in 
temperature of 10° F. for the relative humidity to rise to 70 per 
cent. This humidity is within the danger Umit for the storage of 
material that is subject to staining by damp and mould growths, and 
if allowed to persist damage would be unavoidable. Incidentally, 
this diagram may be used in conjunction wdth the sling hygrometer 
to give an approximate figure for the relative humidity instead of 
using tables. Thus, when the dry-bulb thermometer reads 70° F. and 
the wet bulb reads 60 ° F., the relative humidity wall be seen to He 
between the curved lines 50 per cent, and 60 per cent, and may be 
judged to be about 54 per cent. 

It is interesting to compare continuous records of relative humidity 
and temperature that have been taken out of doors (where the instru- 
ment was screened under standardized conditions) with records taken 
at the same time indoors, and these, in turn, wdth the continuous 
recording from an instrument placed within a closed exhibition case. 
The daily extremes of humidity out-of-doors in London during 
June-September were found to be approximately 30-90 per cent. 
R.H., and these were observed to be reduced m a museum gallery 
to 40-70 per cent. R.H., the mean temperature of the room being 
67° F., whereas, within the case, changes in relative humidity were 
so far reduced as to be almost imperceptible. The comparative results 
are shown diagrammatically in Fig. 2 . The particular exhibition case 






Fig. 2. Summary of records taken concurrently over a period of four m 

London Oune-September). The variations of temperature and relauve huimity m th 
open (broad banck) are to be compared with the less e™ conditions m Ae museum 
g^ry, and variations are still further reduced in an exhibmon case where the record of 
relative humidity is shown to be practically constant. 



used for this test had a capacity of 176 cub. ft., and a glass area of 
88 sq. ft. The effect of the closed case was to slow down atmospheric 
changes until conditions were almost static within the case. The 
fxmction of a glazed case or frame in excluding dirt may be more 
spectacular, but its usefulness in minimizing atmospheric changes is 
no less important when it is a question of long-term conservation. 
This comparative test demonstrates the value of keeping inside a 
closed case or frame objects that are susceptible to moisture changes. 

In a private house, temperature may assume an importance as great 
as that of relative humidity, because collections are usually kept in 
small rooms where the air movement tends to be intensified by the 
opening of doors and windows and by the presence of heating ap- 
phances. In such circumstances there is usually a greater variation of 
temperature than in the larger rooms of a museum, and the use of a 
glass case or frame for susceptible material has imdoubted advantages. 


Temperature and relative humidity have been shown to be inter- 
dependent. In a museum gallery the choice of temperature will be 
pre-determined to suit the comfort of visitors,^ say, within the limits 
of 60-75° F*. and for the collections it will be a question of establish- 
ing hmits of relative humidity that can be considered safe within 
this range of temperature. The limits are determined by observing 
how the most susceptible materials are affected by exposure to 
extreme conditions and to variations of relative humidity. 

A. The tower permissible limit of relative humidity 

Hygroscopic materials are the most sensitive to over-drying be- 
cause they normally contain moisture. Timber that is dried until it 
is in equihbrium with its surroundings is said to be ‘seasoned’ but it 
still contains some 12—15 per cent, of its weight of moisture, and if 

Comfort is of course as much a matter of R.H. as of temperature. In a discussion on 
the effects of the water-vapour content of the atmosphere (MacIntyre, J., Joumat of the 
Inst, of Heating and Ventilating Engineers, 1937, 4, No. 48, p. 570) it was claimed that for 
mental activity an R.H. of under 55 per cent, was ideal, while for rest and recuperation 
something above 55 per cent, and possibly in the region of 70 per cent, was desirable. 



dried below this it will warp. Seasoned timber continues to adjust 
itself in sympathy with changes in the relative humidity of the atmo- 
sphere, and the amount of adjustment will depend on the nature and 
dimensions of the wood, and the variations to which it is exposed. 
Small variations of relative hxmiidity within the range 55-65 per 
cent, are only of significance in the case of thin painted panels where 
the recurrent adjustments may in time cause cleavage between the 
surface of the wood and the paint layer. In the case of heavier struc- 
tiures of wood, significant effects are only produced if the variations 
in relative hiunidity are large and take place quickly. 

The materials other than wood that suffer damage if dried beyond 
a certain point are paper, parchment, leather, and notably the 
adhesives used in making furniture and in bookbinding. Furniture 
that is decorated with marquetry or inlays is particularly hable to be 
damaged through desiccation of the glue, and over-drying has also 
much to do with the physical deterioration of bookbindings. Bind- 
ings soon become dilapidated if they are kept near a radiator or on a 
mantelpiece, or even if exposed for long periods to direct sunhght. 

Taking into consideration the susceptibility of aU organic materials 
to damage by desiccation, the lower safety limit should be fixed at 
50 per cent. R.H. This means that we choose to regard 50 per cent. 
R.H. as the arbitrary Hmit of dryness within the agreed temperature 
range. If, on taking observations, the figure is observed to fall, say, 
to 45 per cent., no great harm would be likely to result, but if such a 
figure persisted over a period, this might well be dangerous, and it 
would be necessary to counter the tendency to over-dryness by 
reducing the heating, or perhaps by increasing ventilation. With 
books or papers either or both methods may be used, but with painted 
panels craclring may be caused by sudden draughts of air. In this case 
it would be safer to increase the relative humidity by cutting down 
the heating. 

B. The upper permissible limit of relative humidity 

The greatest danger that can arise firom an excessively high humi- 
dity is the tendency for moulds to grow on any material that can 



provide nutriment, such as glue, leather, paper. The presence of mould 
growths is a warning that the atmospheric relative humidity is above 
the Hmit of safety. When conditions are favourable to mould growth, 
for example in a Hbrary, a grey dusty bloom is observed in the first 
instance on the darker bindings, and it soon becomes fluffy with a 
tendency to be organized in circular patches. The words mould and 
mildew are used indiscriminately in a popular way to describe growths 
of minute fungi of which there are many different species, though 
superficially they are aU very much alike. They consist of very tiny 
threads called hyphae. The hyphae felt together forming a mycelium 
which throws up fruiting bodies containing spores, and these are 
present in enormous numbers. Fungi thrive under conditions of 
damp, warmth, and darkness, and the materials upon which they 
grow may be stained by contact with the mycehum or with coloured 
matter formed during growth. The growth of moulds can be pre- 
vented by keeping the relative humidity of the atmosphere below 
68 per cent., and this figure is therefore to be regarded as the absolute 
danger limit at temperatures between 60-75° F- hi actual practice 
65 per cent. R.H. is preferable to 68 per cent, for the upper hmit of 
humidity, as this takes account of the fact that papers vary widely in 
hygroscopic nature according to the sizing, filling, &c. 

Closed island cases afford considerable protection from excessive 
humidity. This was demonstrated by a war-time observation. In 
a flooded basement containing ethnographical material, mould 
growths spread apace, whereas in the same basement a closed cup- 
board containing some of the same class of material was found, when 
opened, to be free from mould. As we have already seen, a closed 
case can slow down atmospheric changes, and in this instance al- 
though the cupboard was by no means air-tight, the air inside had 
remained sufficiently dry to preserve the contents. Conversely, it has 
been noticed that when parchment and paper have been packed on a 
wet day, or in a damp locahty, there may be a subsequent growth of 
mildew, arising no doubt from the moist air included in the package. 

The hmits of atmospheric relative humidity are thus defined as 


lying between 50 per cent. R.H. and 65 per cent. R.H. at tempera- 
tures 60-75° F- these conditions are not difficult to maintain in 
temperate climates. Such conditions are satisfactory for museums, 
hbraries, and muniment rooms. Picture galleries are, however, in a 
special category because of the highly sensitive nature of the col- 
lections they contain, and in this case a constant relative humidity of 
58 per cent, at about 63° F. is the ideal to be aimed at. 


It remains, finally, to refer to the problems that arise firom various 
impurities present in the atmosphere. These are regional in character. 

In the first place, there are difficulties pecuhar to the preservation 
of museum materials near the sea-side. Sea air, damp with salt spray, 
may blow well inland, give up its moisture, and become charged 
with a fine dispersion of minute salt crystals; these in turn may serve 
as nuclei for the deposition of moisture, resulting in the formation 
of mists and fogs. In one or other condition, sea air finds its way 
indoors and deposits salt in minute amounts. The salt is hygroscopic 
and maintains a local relative humidity which will support the 
growth of micro-organisms, even in surroimdings that are apparently 
quite dry. This may be a serious matter where there are collections of 
books and papers and, as it is difficult to prevent such contamination, 
it would seem advisable in such locahties to keep things covered as 
far as possible, and to store susceptible material in cupboards or 
glazed cases rather than on open shelves. 

In the second place, there is the problem of damage arising from 
the serious forms of atmospheric pollution in the neighbourhood of 
towns and industrial areas. This is indeed a menace to collections of 
aU kinds and one that is substantially beyond control, for, while soot 
and grit may be excluded, there is no ready means of preventing access 
of sulphurous gases save by the installation of expensive air-condi- 
tioning plant.'^ Sulphurous gases are the cause of widespread deteriora- 
tion. Hydrogen sulphide reacts with all the metals of antiquity (save 
gold) to form dark-coloured sulphides, and is particularly damaging 
to paint films containing white lead (p. 80). Sulphur dioxide is even 



more damaging as it is eventually converted into sulphuric acid which 
attacks a wide variety of materials. The corrosive effects are obvious 
in specimens exposed to the weather — budding stones, or metals — 
but a more subtle and equally damaging form of deterioration due, 
in the first instance, to sulphur dioxide is to be observed in orgamc 
materials which have been exposed to traces of the resvdting acid over 
a long period of time. The rotting of paper (p. 64) and textdes 
(p. 1 12) can often be traced to the action of sulphuric acid, and it is 
this acid, generated perhaps by a local coke stove, that is so often 
responsible for the powdery and decayed condition of old leather 
bookbindings (p. 34). 

AH the sulphur in the atmosphere comes from the burning of 
fuel — coal, coke, or od — and as the sidphur content of these materials 
is knovm, as well as the annual amount of fuel consumed, it has been 
possible to calculate, with what is claimed to be a reasonable degree 
of accuracy, that the annual total of sulphur compoxmds polluting 
the atmosphere in the United Kingdom corresponds to the equi- 
valent of 9 million tons of sidphuric acid — a quantity more than five 
times the annual commercial production of this acid!' Such figures 
serve to emphasize the magnitude of the problem which is almost 
beyond comprehension, but the tragic effects are to be seen every- 

For industrial areas as for sea-side districts the only protection is 
exclusion, as far as possible, of the agencies of decay and this means 
using glazed frames and closed cabinets whether the collections be 
exhibited in a museum or maintained in the more intimate sur- 
roundings of the home. In a closed cabinet the sulphuric acid con- 
tamination of the air will soon be exhausted by its reaction with the 
contents, and no further contamination will take place until the case 
is reopened and a fresh supply of polluted air admitted. 

Such are the main features of the environment as they affect anti- 
quities and works of art. The only way to secure absolute protection 
is by employing methods of hermetic sealing or by using containers 
' The Times, London, 20 April 1953. 



filled with an inert gas (nitrogen or helium), ^ but such procedures 
are only for use in exceptional cases and can hardly be apphed to 
museum objects on a large scale. In general, the most the museum 
curator can do is to recognize the potential dangers inherent in the 
atmosphere, and to counteract them by establishing the best possible 
museum climate for his collections. Where antiquities and works of 
art have aheady suffered damage, treatment to prevent further 
deterioration is essential, and it is hoped that in the pages which 
follow, the curator will find the means to effect this end. 

In Fig. 3 (p. 15) the various factors that contribute to deterioration 
are shown diagrammatically. 


Under the influence of its environment, an object may acquire 
certain features that are of significance in making an assessment of 
its age, authenticity, or provenance. These relate mostly to surface 
effects such as incrustations or weathering, but in the prehminar)^ 
examination the structure and technique should also be considered 
and care taken to select methods of conservation that will not cause 
the loss of any essential characteristics. 

I. Incrustation. Before undertaking any wholesale cleaning of an 
encrusted object it is as well to remember that the incrustation may, 
itself, be of value. It may be a key to the exact find spot; it may 
contain pollen grains that are of potential value for dating purposes, 
or, if the object is a food-vessel, the incrustation may provide interest- 
ing information if the contents can be identified. Records should 
be kept of any details of texture found on muddy incrustations that 
may be evidence of contact with rush mats, basketry, textiles, &c. 
In the event of a number of the same type of objects existing, 

* A Merovingian helmet found at Baldenheim and now in the Strasbourg Museum is 
preserved in a hermetically sealed Perspex case in which air has been replaced by nitrogen 
(France-Lanord, A.,Musee Historique, Lorrain, Nancy). The original documents of the 
Declaration of Independence and of the Constitution have been hermetically sealed in 
hehum-fiUed glass cases and are now on display in the Library of Congress, Washington, 
D.C. Hehum was chosen because it is comparatively simple to detect leaks (National 
Bureau of Standards). 



examples should be set aside with their surface deposits uncleaned, 
as a record of the condition of the objects as found — always pro- 
vided that they can be rehed upon to remain stable under museum 
conditions. Incrustations that are germane to an object and that do 
not conceal inscriptions or ornament should, as a general rule, only 
be removed if they are unstable, or constitute a serious disfigurement, 
and if the object is strong enough to withstand treatment. If an object 
is in fragments, it is usually easier to tell (from the surface staining) 
which pieces fit together before, rather than after, cleaning. 

2. Structure. Examination of the micro-crystalline structure of a 
metal can supply information regarding ancient metal-working 
techniques. It is possible, for example, by a metallographic examina- 
tion to determine whether a metal object has been cast or wrought or 
what heat treatment it has been given in antiquity. To heat the object 
would vitiate the results of any such inquiry, and methods are re- 
stricted, for preference, to those that can be apphed without the risk 
of changing the micro-structure. This is not always possible, how- 
ever, and certainly not in the case of brittle silver objects which can 
only be toughened by a course of heat treatment. 

3. Decoration and technique. Any material that has been fashioned by 
man is hable to bear decoration apphed to or imposed upon its sur- 
face. This may be concealed under incrustations and the greatest care 
is necessary when scraping or chipping or treating with chemicals 
to avoid removing evidence of decoration, and to avoid marking the 
object in such a manner as to lead to possible confusion or misinter- 
pretation of evidence. For the same reason preference should be given 
to the use of modem materials in certain classes of repair work. This 
is important in repairing old textiles or embroideries where, unless 
a synthetic fibre is used, it might be impossible to distinguish old 
threads from new. 



Contaminated Air 




Sulphur Hydrogen Soot Dust 

dioxide sulphide | | 




by desiccation 
Damage to 
of canvas 

I Accident 

Exposure to 
hght, heat, 


Careless handling 
and packing 

Rapid Blackening of lead pigments Pests 

changes Tarnishing of metals I 

Movement of hygroscopic materials 
Warping of wood 
Flaking of paint 
Activation of soluble salts 




Damp Fungi 

Heat Bacteria 

Weakening of adhesives 
Rotting of size 

Staining of paper, vellum, &c. 

Blurring of inks 
Mildewing of leather 

Metallic corrosion encouraged 

Loss of adhesion of iUiuninations 




Wood Beetles 





Adhesion of loaded papers 




Tightening of canvas 








In 1843 Michael Faraday,’ when lectxiring at the Royal Institution in 
London, exhibited leather-bound volumes belonging to the Athe- 
naeum Club that were in a shocking state of decay. This condition he 
attributed to the products of combustion of coal gas, and he pro- 
ceeded to demonstrate that the moisture from a gas flame, condensed 
on cold metal, contained sulphuric acid. The Athenaeum ventilating 
pipes were, in fact, thickly coated with green vitriol (iron sulphate), 
from the action upon iron of the acid fumes evolved on combustion 
and these same acid fumes were responsible for the decay of the 
leather bookbindings and upholstery. 

Faraday’s observations seem to be the earhest on record concerning 
the decay of leather by sulphuric acid, and nothing appears to have 
been done at the time to carry his contribution farther. About the 
turn of the century a committee was appointed by the Royal Society 
of Arts to investigate the cause of leather deterioration, again relating 
to bookbindings: a report was pubhshed, and certain recommenda- 
tions were made, but actually no concentrated attack was made on 
the many practical aspects of the problem until the British Leather 
Manufacturers’ Research Association took the matter up in 1920. 
The Association has made an intensive study of the whole problem 
of the decay and preservation of skin and skin products, and the 
results of twenty-five years’ research have been pubhshed in Progress 
in Leather Science.^ This comprehensive review has been freely used 
in compiling the present chapter. 

' Faraday, Michael, On the Ventilation of Lamp Burners, Royal Institution Lecture, 
7 April 1843. 

^ Progress in Leather Science, ig20-ig4S, British Leather Manufacturers’ Research Asso- 
ciation, Milton Park, Egham, Surrey. 

B 6167 c 2 



Museum collections contain skin products of all kinds, in all stages 
of deterioration, from every part of the globe, and often dating back 
to remote periods. The skins may be plain or ornamented with tool- 
ing or with colour; they may be in the raw condition, or dressed by 
various processes, but whatever their nature it is the task of the 
museum laboratory to take measures for their preservation. Before 
considering what may be done to restore and conserve such objects, 
some attention must be given to the variety of material that is pre- 
sented, and to the methods that have been adopted through the ages 
in preparing animal skins. 


In the earhest times skin had to serve many purposes for which it 
has long since been superseded, and we can gain some idea of its wide 
apphcation by studying our ethnographical collections. The entire 
skin of an animal was sometimes sewn up and used as a water-carrier, 
or filled with air and used as a float or buoy; skin was used to cover 
primitive canoes or coracles; it was used for sails, tents, domestic 
utensils, hunting gear, harness, and accoutrements of every kind. 
The internal organs of animals were also utilized. Vases and water- 
proof clothing were made from the intestines of the cow and walrus, 
floats from bladders, and utensils from the stomach of the camel. 
Masks were made by moulding moist skins or bladders over a suit- 
ably modelled shape and allowing them to dry in position; the shape 
of the mould was retained by the tissue, which could then be stabi- 
hzed by waterproofing with fat or oil. 

Untanned skins and skin products are, however, not permanent 
materials; they are hable to attack by micro-organisms, and are very 
sensitive to moisture. These weaknesses can be overcome to a certain 
degree by methods of curing and dressing, and there is no doubt that 
such methods were employed in primitive times to protect the skins. 
Even at a time as remote as the last interglacial period* there is evidence 
to suggest that simple methods of skin-dressing (smoking and treat- 
ment with oil) may have been practised with the object of preventing 
‘ Ganiser, A., ‘Early History ofTanning’, CIBA Review, 8, p. i. 

Flakinu; due to crystallization ol salts on the surtace 


{Coiirrc:>y ot Biifi^li Lcii'lici RcA\in!i 



(t) Physical Decay — {Lcjt Pair) — Hinges dry and cracked. 

( 2 ) Chemical Decay — (Centre Pair ) — General deterinration to powdery red surface. 

( 3 ) Use ot 13. M. Leather Dressing — {Riahr Pair ) — Books were in the condition ol (t) 
betore treatment. 



putrefaction and enhancing flexibility. However, skin does not last 
long in damp earth, and no actual examples from this age are still 

It is otherwise in the dry and sterile tombs of Egypt where many 
specimens of ancient skin have survived from Predynastic times. 
These are rigid and brittle as a result of desiccation, and it is, there- 
fore, not an easy matter to determine whether they have been cured 
or processed. There is ample evidence that skin was exploited in 
ancient Egypt for utihtarian as well as for decorative purposes. In 
Tut-ankhamun’s tomb it was used embedded in gesso to form a 
resflient backing for gold leaf that was later ornamented with tooling, 
and it was even used by the Egyptian scribes in the form of a scroll 
for writing upon (p. 39). T hi s latter use need occasion no surprise, 
as the flesh side when scraped and cleaned yields a smooth surface of 
pleasant colour which ‘takes’ ink satisfactorily. In this respect, how- 
ever, it falls far short of parchment, both as regards its suitabflity for 
the purpose and its lasting properties. 

Parchment did not appear until the second century B.C., and then 
not in Egypt but in Pergamon (Asia Minor), from which its name 
is derived. The material is made from the tough, white corium, or 
inner layer of animal skin, by a process that relaxes and flattens the 
capillaries so that they are unable to reabsorb moisture. Though it is 
easily stained, and deformed by heat and damp, it is not subject to 
chemical deterioration, and has proved to be extraordinarily durable 
as a material for fine writing. Parchment is dealt with in detail 
on pp. 44 seq. 


The word ‘skin’ is used as a general term for aU classes of material, 
whether raw, cured, or processed, and skin from the larger animals, 
such as horse and cow, is known specifically as ‘hide’. ‘Curing refers 
to a first aid or field treatment to prevent putrefaction, and pro- 
cessing’ to any more permanent treatment, including the manu- 
facture of leather. 

The science of the preservation of leather begins with the study of 



the untanned skin — the raw material before it is processed — and it 
may contribute to an understanding of the problem if we consider 
briefly the main physical features and the micro-structure of animal 

When an animal is skinned the pelt is seen to take the form of a 
continuous membrane of fibrous tissue, having a hair side and a flesh 
side, the hair side being patterned with sweat glands and hair foUicles 
(grain), and the flesh side hned -with fat, muscles, and blood-vessels 
(PL 2). Skins from the different animal species exhibit great varia- 
tions in physical structme, and this is most pronounced as between 
warm-blooded and cold-blooded animals. Even in the same species, 
however, where the pattern or grain tends to be constant, variation 
in micro-structure may be observed, and is caused by such factors as 
the age of the animal and quahty of feeding. Some skins are valued 
primarily for their surface covering of hair, fur, or wool, some for 
the skin itself (rather than for the surface covering), whereas others, 
notably in the arctic regions, serve both purposes, the skin being 
impervious to the elements, and the fur acting as an insulator and a 
conserver of heat. 

In the micro-structure of animal skin the true skin tissue or corium 
is composed of a reticulated network of protein fibres, the main 
constituent being the protein knovra as coUagen — ‘the glue producer’. 
(If skin is boiled in water for a sufficient time, it undergoes hydrolysis 
and is converted into glue.) Recent chemical research has shown that 
collagen has a chain structure, i.e. it consists essentially of long mole- 
cules composed of many atoms attached together like links in a chain, 
with smaller side chains branching from the main backbone. These 
comphcated molecules of collagen tend to orient themselves in the 
same direction, to associate in bundles, and to retain water perti- 
naciously in a loose form of chemical combination. 


In the normal course of events, the skin of a slaughtered animal is 
speedily a prey to the growth of moulds and bacteria, but the mechani- 
cal removal of associated fats, muscles, and blood-vessels e limin ates 



much nutrient material, and the drying of the skin goes a long way 
towards rendering the tissue proof against putrefaction. In this 
condition, however, the skin is of httle practical value; when the 
collagen fibres have been deprived of their water, the skin becomes 
homy and brittle, and flexibUity can only be restored by relaxing the 
fibrous bundles so that they sUde freely over each other again, as they 
did before the skin was dried. This relaxing of the skin may be ac- 
comphshed in several ways, notably by some form of prolonged 
manipulation, or by the incorporation of lubricants, or, usually, by a 
combination of both methods. 

The tramping of skins in tubs has been a traditional method of 
preparation in some parts of the world for many years. The Eskimos 
soften skins by chewing, a process which doubtless is enzymatic in 
action as well as being mechanical. Manipulative processes may, in 
themselves, be sufficient, but generally improved results are obtained 
by the use of dressings, such as castor oil, or sulphonated neat’s-foot 
oil. These are often appHed in emulsified form, and they penetrate 
and lubricate the tissue, partially replacing moisture as the skin dries. 

Fur skins 

An animal skin ‘in the hair’ (or wool) prepared for the taxidermist 
may have been softened by oil dressing, or merely worked mechani- 
cally to a condition of relative flexibUity. Equally suitable as a method 
of preparing soft-dressed fur skins is the following process. The fresh 
skin is stretched out fur downwards and pinned to a board; the 
flesh side is scraped with a knife having a convex blade, until the sur- 
face of the skin is approximately uniform, and then the skin is 
allowed to dry. It is wetted again, and the process repeated. There 
seems to be a positive advantage in alternate drying and wetting, but 
as it is important in the preparation of fur skins to prevent putre- 
faction, the sequence of operations must not be interrupted; a httle 
carbohc acid may be used if required to keep the skin sweet, or 
alternatively some salt and alum. Finally, the skin may be pulled 
over a beam-knife or stake. In the hands of an expert this yields a 
very soft and flexible product. 



Alum dressing 

The mention of alum suggests a process of skin dressing called 
tawing, which dates back to remote antiquity. This consists in treating 
the skin, after prehminary cleaning, with excess of alum, the resulting 
product being a substance light in colour, and having the properties 
of leather so long as the alum is present. The action is reversible, how- 
ever, and if the alum is removed by washing, as is easily possible, the 
skin is ruined. Water must not be used, therefore, with tawed skins. 
Tawing is still practised today in the manufacture of wool skins; 
tawed leather is used for making gloves, and to a limited extent for 

Oil dressing 

Oil tannage is employed today in the preparation of chamois 
leather. For this purpose the skin is spht on a band-knife machine, 
the grain-spHt or surface side of the skin (skiver) being worked up 
as leather for bookbinding, while the flesh side is drummed with 
cod oil and thus converted into chamois. In chamois leather the fibres 
are protected with oxidized oil, and the skin absorbs water readily 
as there are comparatively wide capillaries between the fibres. Alter- 
natively, the flesh side may be converted into parchment (p. 45). 


Leather is animal skin which has been dehaired, defatted, neutral- 
ized, and rendered non-putrescent by a form of treatment that makes 
the skin impervious to water, while preserving the natural phabflity. 
This treatment is carried out with a tanning agent which may be of 
vegetable or mineral origin. 

Vegetable-tanned leather. One of the earhest materials used for tan- 
ning leather in northern countries was an infusion of oak bark — the 
Breton word tann is the equivalent of oak. This infusion contains 
tannins which are chemical substances that have the property of com- 
bining with the collagen and other protein fibres of the skin, replac- 
ing the loosely boimd water, and reinforcing the side chains or cross 



links of the protein molecules. The result is a material having qualities 
greatly superior to those of raw hide, particularly as regards water 
resistivity, durabihty, and what is known as ‘feel’ — an elusive quahty, 
but one that is perhaps sufficiently self-explanatory. 

In ancient Egypt it is probable that tanning was first carried out 
by the use of acacia pods, rather than oak bark, and. Lucas^ mentions 
that articles of leather are common in Tasian, Badarian, and Pre- 
dynastic tombs. The Egyptian tomb paintings provide some interest- 
ing illustrations of contemporar}- methods of leather-working, 
notably in a mural at Gumah (seventeenth dynasty), and also in the 
tomb of Aba (No. 36) in the Asasif at Thebes (twenty-sixth dynasty). 
Some of the processes seem to differ fittle from the methods em- 
ployed at the present day. 

Besides oak bark and acacia pods there are many other plant pro- 
ducts widely distributed in nature that have the property of stabi- 
li2ing the protein structure of skin, and giving it the properties of 
leather. These vegetable tannins may be classified in two chemical 
groups, the catechol group, and the pyrogaUol group, though certain 
of them — and oak bark is an example — ^belong to both categories. 
These chemical groups are of significance and wiU be referred to 
later, in studying the permanence of leathers used in bookbinding 
and upholstery. 

Mineral-tanned leather. Alum tawing may be regarded as a form of 
mineral tanning, but this description is generally reserved for the 
modem process carried out by treating the skin with a salt of chro- 
mium which reinforces the side-chains of the collagen fibres as 
efficiently as the vegetable tanning agents, to produce leather of a dis- 
tinctive quahty. Chrome leather is practically non-wettable and very 
durable both physically and also chemically, since it is not so vulner- 
able to attack by sulphuric acid as vegetable-tanned leather. 

If the two types are compared as potential bookbinding leathers, 
the main difference lies in the fact that vegetable-tanned leathers can 
be relaxed and manipulated in the damp condition, and moulded to 
the shape of the book, whereas chrome leathers are intractable and 
’ Lucas, A., Ancient Egyptian Materials and Industries, 3rd edition (revised 1948). 



difficult to relax and to mould permanently to any required shape. 
Vegetable-tanned leathers take gold leaf and stamped decoration or 
tooling well; chrome leathers badly, or not at all. Chromes, on the 
other hand, are pre-eminent in their resistance to attack by moulds 
and chemicals. One day, perhaps, by some form of combined vege- 
table and mineral tanning, it may be possible to produce a new kind 
of bookbinding leather of improved quahty and durability, in which 
the characteristic weaknesses of the chrome and vegetable-tanned 
leathers are suppressed, and the virtues of both preserved. Meantime, 
we are faced with the problem of preserving objects of skin and 
leather that differ widely in stabihty. 



Moisture. In extreme cases the degradation of leather by moisture 
may proceed to the Umit, and nothing may remain in an excavation 
to indicate that leather was once present, save, perhaps, indefinite 
traces in the sand, such as were found at Sutton Hoo between the 
various pieces of the purse complex. Gold coins were lying loose in 
the sand beneath the metal ornaments of the purse flap, and these no 
doubt were originally in a leather bag. Belt fasteners were also found 
in the sand with only a stain to mark where the leather belt had once 
been lying. That bacterial action contributes to such degradation is 
beyond question, and vestiges of leather sometimes remain as recog- 
nizable tissue where they have been stained green with the products 
of corroded copper which have acted as a sterflizing agent. Peaty 
waters and very salt waters also have a protective effect. Exposure 
to the prolonged action of a low humidity, as in some Egyptian 
tombs, has converted skin into a black syrup of bitumastic ap- 
pearance, which is sometimes found to be still tacky, but sometimes 
has run into a hard black sohd with a superficial resemblance to 
ebonite. Both the syrup and the sohd material are soluble in water. 
Leather manuscripts have been found that are only converted in part 
to this black material (see p. 40), and there is then the problem of 



salvaging what remains before the writing is obhterated due to stain- 
ing by the syrupy decomposition products. 

Fungus. In ethnographical collections and in Ubraries, and, indeed, 
wherever skin and skin products are stored, humidity control is most 
important, because moulds grow readily on these materials when the 
relative humidity of the atmosphere is above 68 per cent, at ordinary 
temperatures. The mould growths cause staining, and erosion may 
take place as well. Dyed leathers may change colour as a result of 
fungus attack, and leather bindings are thus particularly prone to 
damage. The use of fungicides to prevent the growth of mould is 
merely palliative, and is not as a rule advocated. The real cause of 
the trouble is the condition of dampness in the collection, which 
would be deleterious to all organic materials, and it is of first im- 
portance to take steps to restore dry conditions. Good ventilation is 
essential, and will speedily arrest the spread of mould growths. In 
countries where there is persistent high humidity, however, the use 
of fungicides is recognized as a necessity. 

A suitable fungicide for appUcation to leather is one which does 
not cause staining or rotting, and is of low volatihty. Paranitrophenol, 
even in moderate concentrations, causes a yellow stain, but it has 
been proved to be a rehable fungicide for leather when employed in 
exceedingly dilute aqueous or alcohoHc solution, strength 0-35 per 
cent.; and pentachlorophenol is effective at 0-25 per cent, strength. 
When these tw'o fungicides are mixed in equal parts by weight, it 
has been found by the National Bureau of Standards, U.S.A.,' that 
they can be incorporated in combined concentrations as low as 0-2 
per cent, and still be effective for the preservation of leather against 
fungoid attack. Such a mixture was an official issue to the American 
Army in the tropics towards the close of the Second World War, 
and proved to be highly satisfactory. When it is desired to afford pro- 
tection over a prolonged period, greater concentrations of fungicide 
are, in general, necessary, and it is no longer possible to use paramtro- 
phenol on account of its staining properties. Preference may then 

’ Microbiological Deterioration of Organic Materials (Mis. Pub. 188), 1948, p. 36. U.S. 
National Bureau of Standards, Govt. Printing Office, Washington 25, D.C. 



be given to derivatives of pentachlorophenol. Thus, the sodium salt 
known commercially as Santobrite' can be dissolved either in alcohol 
or in water to make a 2 per cent, solution that is useful for impregna- 
tion; or, should a spraying technique be preferred, the lauryl deriva- 
tive may be selected, a convenient preparation being the paraffin 
solution available commercially as Mystox L.P.^ 


It is necessary to take special measures to protect skin products 
firom attack by insects, and even imder the best conditions skins 
should be inspected regularly and cleaned firom time to time. Furs 
are especially prone to attack by moth larvae. Leather bindings too 
are hable to be damaged by moth, as well as by several other insect 
parasites in northern countries, notably those of the order Coleoptera. 

In a museum provided with special equipment, the usual first-aid 
treatment for an infestation on a large scale is to fumigate with 
hydrogen cyanide, methyl bromide, or carbon disulphide. These 
chemicals are all effective, but none confers lasting protection, and 
it is essential to examine the store and, if necessary, to treat the room 
also to ensure that reinfestation will not immediately take place. 

To deal with small outbreaks, spraying is generally adopted, and 
a good hand-operated spray-gun can be very effective for this pur- 
pose. The insecticide should be highly atomized and brought into 
contact with the tissue. Packages must therefore be opened and 
folded leather and furs unwrapped, otherwise the labour will be 
expended in vain. The spray deposits a very fine film of chemicals 
that retains its insecticidal properties for some considerable time, and, 
from this point of view, spraying has an advantage over fumigation. 
Where spraying has to be done regularly on a large scale, a power 
spray may be regarded as a necessity, and this is generally employed 
at a pressure of about 30 lb. per sq. in. 

An insecticide for the control of moth is usually dissolved in an 

' Monsanto Chemicals Ltd., Victoria Station House, London, S.W.i. 

^ Catomance Ltd., Welwyn Garden City, Herts. 

(Hinges broken— covers detached— leather scrutfed- headbands torn) 





Onlv protected leather, {bottom senes and No. lo) have been able to withstand 
this accelerated ageing test. (See Appendixes VIll and IX) 



odourless paraffin distillate that can be guaranteed not to cause stain- 
ing, and the spray may act on the parasites either as a stomach poison, 
or a contact poison, or both. It is a good idea to vary the insecticide 
occasionally, as insects have been found to build up a resistance to a 
particular poison. This has been observed in the case of DDT. 
Though slow in action DDT is a very effective insecticide, parti- 
cularly when used with pyrethrum, which is immediately lethal to 
moth, and is said to have a high ‘knock-down’ value. Many effective 
commercial preparations are available, but if a fair quantity is 
required it is cheaper in the end and more satisfactory to prepare the 
solutions oneself than to use preparations of unknown composition. 

The following formulae can be recommended. 

1. Pyrethrum I DDT spray solution. Dissolve i oz. of pyrethrum con- 
centrate (Pyefly)^ and i oz. of DDT powder^ in one pint of odour- 
less distillate. 3 

2. Lethane spray solution. Dissolve i oz. of deodorized Lethane 
384^ in pints of odourless distillate. 

Fur skins used to be prepared as a routine by rubbing with arseni- 
cal soap, but this has given place more recently to treatment with 
boric acid, or with borax, which is claimed to afford just as effective 
protection against insect attack as the arsenical compound, without 
the danger to the human subject attendant on the use of arsenic. 

Where skin specimens are known to be free from insects, it is a 
wise precaution to shut them away or isolate them in tissue paper or 
in Cellophane. This keeps the objects free firom dust and may prevent 
parasites getting at them. A Cellophane hanging wardrobe as used for 
textiles may sometimes be brought into service for skins, and the 
Polythene bags provided with zip-fasteners are also very useful. 
These may be charged with a handful of dichlorobenzene cr\'stals. 
This also appUes in the case of susceptible material kept in drawers, 

‘ Pyefly obtainable from Messrs. Stafford, Allen & Sons, Ltd., Cowper Street, Fins- 
bury, E.C.2. 

^ DDT obtainable from The British Drug Houses Ltd., Poole, Dorset. 

^ Odourless distillate obtainable from Messrs. Shell Mex & B.P., Shell Mex House, 
Strand, W.C.2. 

Lethane 384 obtainable from Messrs. Lenmg& Co,, 177 West Street, Erith, Kent. 



cases, &c. The insecticide is effective so long as some crystals remain 

Insects in books 

Many insects damage books by tunnelling through the bindings 
or leaves, and, when infestation is general, wholesale fumigation, e.g. 
by hydrocyanic gas or carbon disulphide, may be necessairy in order 
to eradicate them. 

When only a few books are badly infected, these may be exposed 
for about a month to the vapours from dichlorobenzene crystals 
(about 2 oz. per cub. ft. of air space). The books should be opened 
fan-wise and placed, with the crystals, in a small air-tight box, e.g. a 
biscmt tin sealed with Sellotape. A loose paper lining is inserted to 
prevent the books coming in contact with the tin or the insecticide. 

Often, however, the attack is less severe, and can be controlled by 
thoroughly cleaning the books and bookcase, wiping over the 
shelves with a rag moistened with a hquid insecticide, or spraying 
the insecticide as a fine mist from an atomizer. Isolated attacks may be 
checked by applying Hquid insecticide to the larva holes by means of 
a fountain-pen filler or water-colour brush. Bindings may be sprayed, 
but, as some papers are Hable to be stained, the paper should not be 
sprayed. As insecticides do not necessarily Idll eggs, the bindings 
should be inspected later and treatment repeated if required. Either 
of the spray solutions mentioned above (pyrethrum or Lethane) is 
satisfactory for this purpose, and they can be reHed upon to kill moth 
and larvae, and to render the foodstuffs distasteful to all the common 
book pests. If a spray is used, some prehminary practice may be 
necessary with the atomizer in order to ensure that a mist and not 
a stream of Hquid is deHvered from the orifice. 


For the preservation of skin and leather in the museum coUections 
it is fortunately not essential to be able to identify the method of 
processing. Raw uncured skin, when damp, putrefies. If kept in a 



reasonably dry atmosphere, skins, whether raw or cured, even if they 
have been flexible, tend in time to revert to the condition described 
as homy. Leathers are aflected to a less degree and vary according to 
type, but there is generally a cumulative hardening. 

It is no uncommon occurrence to find skins neady folded in parcels 
where they have become rigid owing to years of neglect. To attempt 
to unwrap the brittle skin without some form of preliminary treat- 
ment would be to court disaster. If cracking has not already taken 
place, such specimens can usually be recovered in an undamaged 
condition by going over the skin, and particularly the folds, with 
a damp sponge, following this with a leather dressing. The moisture 
helps the dressing to spread easily. It should be given a Httle time to 
penetrate, and then it will be possible to work the skin gradually 
back into its original shape with the fingers. Skin clothing of all 
kinds should preferably be kept on coat-hangers, where it is less 
likely to suffer than if wrapped up and put away ia drawers. 

When skins are damaged by vrater, there are invariably comphca- 
tions due to the inevitable fluffy growths of moulds that carpet the 
affected specimens, concealing ornament and disguising colour. A 
case in point was provided by a skin cloak (carossa) from Angola, 
which emitted an unbearable odour of putrescence, and was usually 
hidden under a covering of green mould. Preliminary cleaning was 
done in the open, using a soft dry brush. This removed the mould 
and revealed the clammy skin, which was purphsh-red in colour, 
due to the use of the primitive dye-stuff, cam-wood. The cloak was 
then pinned out in a shallow case for exhibition purposes, and the 
atmosphere in the case was gradually desiccated by means of dry 
sihca gel (Appendix XI). After changing the sihca gel several times, 
a paper hygrometer that had been placed in the case recorded a 
figure of 60 per cent, relative humidity, and as this is a perfectly 
satisfactory figure for the preservation of organic materials susceptible 
to mould attack, it was unnecessary to treat the cloak with fungicides. 
The specimen has since given no further trouble. 

Leather objects are occasionally found in bogs or damp ground, 
and these require special treatment. Though greatly weakened and 



distorted, the leather may retain a certain flexibiHty in the wet condi- 
tion, which it would lose permanently if allowed to dry. In no 
circumstances should formahn be used on specimens of this kind as 
it makes the leather rigid. The object should first be photographed 
and measured, and the surface then washed with water, using a fine 
brush. It may then be rubbed over with a 2 per cent, solution of 
carboHc acid in alcohol; the leather should be immersed immediately 
afterwards in molten vaseline, and kept there for a day or longer at 
a temperature of about 80-100° C. This improves the appearance, 
and in favourable circumstances softens the leather still further so that 
it may be moulded back to shape. But there is usually a fair amount 
of shrinkage, and as the vaseline leaves a sticky surface, it is generally 
desirable, having recovered the shape, to replace the vaseline with 
paraftin wax. This is done by immersion for half an hour in molten 
wax at 1 10° C. The leather is then removed, and may be stuffed with 
soft paper to maintain the shape as it cools. A very Httle bitumen 
powder may be added to the wax to darken it, so that it will not 
show as a light film on the leather. 


When leather has to be repaired this is best done with a paste 
mixture^ before the appHcation of leather dressing, because, once the 
tissue has been lubricated with an oil dressing, the paste cannot be 
made to adhere. When degreasing is necessary prior to repair, this 
can be accompHshed by trichlorethylene, but it should be applied 
locally to the parts that have to be stuck together, and sparingly, or 
it may remove tanning material. Leather dressing may be appHed, 
of course, after the repair has been carried out. 

When leather is rotten and brittle it is impossible to put new life 
into it, but even if the surface is cracked or broken it may be possible 
to strengthen it by sticking a strong material to the bacL Such 
restorations are carried out in the case of flat leathers — screens, 

‘ Grade No. 753, a cement made from paste and rubber and supplied by National 
Adhesives Ltd., Slough, Bucks., is ideal. It may require to be t hinn ed with water a htde 
before use. Flour paste (see Appendix III) is only suitable for thin leathers. 



leather wall-coverings, and the flat parts of leather upholstery. If 
the surface of the leather is very dirty, it should first be cleaned by 
sponging it over with a httle size water. ‘ The leather is then re- 
moved (if necessary) from the object to which it is attached, and 
laid face downwards on a table covered with glass or linoleum, and 
if it is wriakled, an attempt should be made to flatten the skin by 
damping, allowing time for the moisture to penetrate, and then 
gradually lowering a piece of plate glass upon it. When the leather is 
well relaxed, pressure may be apphed by placing suitable weights on 
the glass and the pressure should be retained until the skin is dry. 
A piece of raw cloth canvas of suitable size and weight is now pasted 
to the back of the leather, the canvas is covered with newspaper, and 
pressed overnight. Next day the leather will be found to be consider- 
ably strengthened. 

This will be all that is required in the case of a wall-covering. A 
screen panel would be remounted, possibly after applying webbing 
to the framework to prevent the leather being damaged if the 
screen were to be carelessly handled. The side of a leather chair is 
sometimes backed with thin tough miUboard prior to remounting 
the leather. 

It does not follow that because an object is of skin or leather it 
should necessarily be treated with leather dressing. It may be desir- 
able to keep rigid leather stiff, rather than to make it phable, and 
caution is required in dealing with tooled leather, as the sharpness 
of the pattern might be impaired were the leather to be softened. 
These remarks apply to such things as wooden chests covered with 
hide, and to accoutrements, scabbards, &c. Circumstances must 
determine what is the best treatment, whether it is better to con- 
sohdate by using some form of thin lacquer or microcrystalhne wax 
(see Appendix XII), and so restore the appearance, or whether to aim 
at restoring flexibflity. 

' On no account should commercial detergent powders be used for cleaning leather as 
they are liable to extract the tanning materials and ruin leather and chamois. Castile 
soap, soft soap, and saddle soap are harmless to leather provided they are used in strict 
moderation, but they may affect dye-stuffs. 

B 6157 





Vegetable-tanned leather, the only kind of leather that we have 
to consider in this section, is subject to two forms of deterioration, 
physical and chemical, and these should be distinguished, though both 
may be present together in one and the same binding. Physical 
deterioration is due to normal wear and tear, and commonly takes 
the form of a breakdown of the surface, exposing the tough fibrous 
underlayers of the skin which are not easily scruffed with the finger- 
nail. Chemical deterioration is accompanied by cracking and often 
by change of colour, and sooner or later it involves a complete 
breakdown of structure. Leather in this condition may be reduced 
to a fine powder by scruffing. Chemical decay takes place more 
quickly than physical decay, and its effects are more profound. Pi. 3 
shows a series of leather bindings in which the hinges have suffered 
decay, partly physical and partly chemical in nature, since the period 

Chemical deterioration. When leather volumes are standing together 
on a book-shelf the sides are protected, but the backs are exposed, as 
well as the top headbands, and {especially if in an industrial area), 
the exposed parts absorb sulphur dioxide from the atmosphere. 
Leather bindings invariably contain traces of iron compounds; these 
act as a catalyst in the conversion of sulphur dioxide to sulphiu: 
trioxide which, in turn, unites with water to form sulphuric acid. 
The sulphuric acid attacks the leather tissue in presence of oxygen, 
causing it to become very brittle. The opening of the books cracks 
the joints at the top, and this is usually the first visible sign of chemical 
deterioration. It is only the beginning, however; the hinges eventually 
become so weak that the covers spht off, and the back may eventually 
fall to pieces (see Pi. 4). 

In studying the deterioration of leather in a long series of dated 
bindings, such as the heavy volumes of The Times newspaper, bound 
in pigskin at the British Museum, certain anomahes can be observed. 
Some of the original bindings are still in use today after a hundred 
years’ service — they are worn and frayed round the edges, as is only 



to be expected, but they show no signs of the powdery chemical 
decay that occurs on similar volumes later in the series. The younger 
leathers have decayed, while the older ones have survived. The 
second example is that of the General Catalogue of Printed Books 
(‘G.K. i’), which is in daily use at Bloomsbury. It contains service- 
able though badly worn volumes of about fifty years of age in their 
original bindings of green straight-grained morocco, whereas others 
in the same series of only about fifteen years of age have had to be 
repaired, and some of the volumes have actually had to be rebound. 

As a result of investigations initiated by the British Leather Manu- 
facturers’ Research Association, and carried out by R. Faraday Innes,i 
it has been found that it is the presence of certain water-soluble sub- 
stances in the old leathers that has protected them firom chemical 
attack. In the past, leathers used to be coloured by surface-staining, 
but modem processes involve immersion of the leather in a dye-bath. 
No doubt colour-matching has been facihtated by the modem 
methods, but immersion of the leather means that the protective 
water-soluble substances have been extracted, and the lasting quahty 
of the leather has thereby been impaired. Innes carried his researches 
a good deal farther, and showed that if the water-soluble substances, 
‘non-tans’, have been removed from freshly manufactured leather, 
it is possible to introduce certain simple salts which wdl confer pro- 
tection (Appendix IX). He also devised an accelerated form of test 
called the ‘Peroxide’ or 'P.I.R.A.’ test (Appendix VIII), that will 
indicate whether a given sample of leather is likely to be permanent, 
i.e. whether the leather will resist deterioration when it is exposed to 
oxidation in the presence of sulphuric acid. 

The results of applying the peroxide test to a series of bookbmding 
leathers are shown in Pi. 5. The first row contains ‘controls’ set 
aside for comparative purposes — they all seem to be sound leathers. 
The second row shows the result of applying the peroxide test to 
other samples of the same leathers — only No. 10 has survived. The 
third row gives results similar to the second; in this case the specimens 
were washed before applying the test, but as the leathers had already 
' British Leather Mfrs.’ Res. Assoc., Laboratory Reports, 1933. 12, 228. 



been deprived of their ‘non-tans’ during manufacture, the washing 
made no difference. In the fourth row the leathers were protected, 
before testing, by potassium lactate in accordance with the method 
described in the appendix. The photographs show that the treatment 
recommended makes it possible for the leathers to survive intensive 
oxidation in the presence of sulphuric acid and it is thus possible to 
ensure the hfe of vegetable-tanned leather by lactate treatment. 

‘Real leather’ can only be rehable for bookbinding if it is ade- 
quately protected by the presence of non-tans or salts. To ascribe any 
virtue to ‘acid free leather’ is purely illusory, as leather will absorb 
sulphur dioxide in any case in polluted atmospheres. 

It should be added that protective salts are only of value when 
apphed to freshly tanned leather: they have an insignificant pro- 
tective action on old leather that has aheady absorbed some acid and 
is suffering chemical decay. 

The problem of protecting leather against chemical deterioration 
is stiU receiving attention. There is prospect that it may be solved by 
the apphcation of reagents to deprive the iron in the leather of its 
power to catalyse the formation of sulphuric acid. This has been 
shown to be possible by using a water-soluble oxalate,^ and also by 
sequestrating the iron (see p. 278) with potassium pyrophosphate. ^ 
The efficacy of these methods can only be determined by the test 
of time, however, as in this case the peroxide test is not apphcable. 
This idea of sequestrating the iron is still only of laboratory interest; 
so far as is known no commercial book-binding leathers are yet avail- 
able that are claimed to be protected in this way. 

A long-term test. In order to test the various factors that might be 
supposed to have a bearing on the permanence of vegetable-tanned 
leather, two identical series of bindings have been prepared, one series 
being kept in the clean atmosphere of the National Library of Wales, 
Aberystwyth, and the other set in the Department of Printed Books 
at the British Museum. The bindings have been done in leather from 
different animals, various tannages are represented (sumac, gambier, 

‘ Cheshire, A.,Journal of the International Society of Leather Trades Chemists, 1946, 30, 134. 

^ Innes recommends the use of a i per cent, aqueous solution. 


oak bark, &c.), and the leathers are either unprotected or else protected 
indifferent ways by impregnating with a variety of salts. Some leathers 
have been degreased, some treated with leather dressings, &c., the aim 
being to make the series as representative of aU the variants as possible. 

Each series contains over 600 volumes. These are examined every 
few years and, although the scheme has only been in operation for 
twenty years, some interesting information has already been ac- 
cumulated.’^ Some of the main conclusions may be here enumerated: 

1. The volumes at the British Museum show considerably more 
deterioration than those at Aberystwyth, where the atmosphere is 
practically free from sulphur dioxide. 

2. Salt-free tannages of the pyrogaUol type are more resistant than 
those of the catechol type, and it is beginning to be possible to dis- 
tinguish which actual tannages are best and which are least durable; 
thus, sumac and Nigerian acacia pod tannages are the best of the 
pyrogaUol type, and gambler the least durable of the catechol type. 

3. Removal of soluble non-tans facihtates decay. Addition of 
certain salts increases the resistance even of catechol-tanned leather. 

4. The peroxide test is a rough guide as to how durable vegetable- 
tanned leathers are likely to be in poUuted atmospheres. 

5. The foUowing are not major factors in the resistance of bindings 
to chemical decay; type of skin, type of cure, presence or absence of 
grease, method of hming and pickhng. 

This short account of the scientific work that is being done on 
vegetable-tanned leather brings to notice three matters of practical 
importance m the preservation of bookbindings and of leather up- 
holstery. In the first place, these leathers wiU be inclined to deteriorate 
most quickly in rooms where the air is poUuted by the fumes of 
combustion of gas, coal, and especiaUy of coke. Secondly, washing 
the leather may be equaUy bad, in that it wiU tend to remove soluble 
non-tans, and predispose the tissue to attack. Thirdly, leather dress- 

’ British Leather Mfrs.’ Res. Assoc. 1932, ii, 276. Long-term test commenced. 1945, 24, 
104. 1st Interim Report. 1952, 31, 380. 2nd Interim Report (limes, R. F., Plenderleith, 
H. J., and Moss, A. A.). 1954, 33, 208. 3rd Interim Report (limes, R. F., Plenderleith, H. J., 
and Werner, A. E.). 



ings have httle effect in protecting leather from chemical decay, how- 
ever useful they are in the physical sense in lubricating the tissue and 
thus keeping the leather supple. 

Physical deterioration. Physical deterioration is determined in no 
small degree by bindery practice, and can be controlled to a large 
extent by the selection of suitable leathers for different sizes and 
weights of books, as well as by handling. Sharp comers on metal 
shelving are an obvious cause of fraying .When books are too tightly 
packed on the shelf the bindings become deformed, and the effort to 
pull a book out may result in the top edge of the back being tom. 
The evil practice of building shelving over radiators results in the 
speedy deterioration of books kept thereon; heat causes leather to 
harden and glue to become desiccated; and, if the shelving is open, 
a dust problem is added to the others. 

Leather dressings for library use. In the absence of chemical decay 
the hfe of a book depends largely on the condition of the hinges. 
These should not be allowed to become hard and brittle, and this may 
be prevented by the occasional appHcation of a leather dressing. The 
most suitable form of leather dressing for hbrary use is considered 
to be a mixture compounded on a basis of lanohn; the lanolin sinks 
into the tissue in the course of a few days and lubricates the fibres. 
The mixture should also contain a httle wax; wax does not sink in, 
but remains on the surface to consohdate any powdery leather, and 
it may be poHshed in a day or two with a soft cloth. The dress- 
ing should be a hquid so that it can be apphed without strain, and 
it should be used sparingly, as an excess of grease would be likely 
to stain hght leathers and possibly even the pages of the book. For 
the composition of the British Museum Leather Dressing, see Appen- 
dix X. Similar preparations are now available commercially. ^ The 
occasional use of such a leather dressing wih prolong the hfe of a 
leather binding whether or not it is suffering from chemical decay, but 
there is no way of stopping chemical decay when this has once set in. 

Washing books. When books are in constant use the bindings tend 

* e.g. ‘Pliantine’; obtainable from Arthur Rich & Partners, Ltd., Factory A.2, Treforest 
Trading Estate, Pontypridd, Glam., Wales. 



to become very dirty, and it may be desirable to wash them before 
applying leather dressing. But washing vegetable-tanned leather is 
only permissible when protective salts are added thereafter. The pro- 
cedure is as follows. The books are sponged Hghtly with water con- 
taining a httle Castile soap. This is removed with a clean damp 
sponge, and then the books are set up to dry. They are next sponged 
with potassium lactate (lo per cent.). Appendix IX, and left over- 
night. The leather dressing is appHed next day, and after a suitable 
time rubbed up with a soft cloth. In cases where books are washed 
annually, the strength of the lactate may be reduced to 5 per cent. 
Potassium lactate solution should be freshly prepared and not left 
lying about exposed to the air, as it tends to go mouldy on keeping. 


Two cases involving the salvage of documents of unusual interest, 
written on leather or skin which had long since decayed, are here cited, 
because each involves a principle which is, in its way, a novelty. 

The first example is a Hieratic leather roll dating £com the Hyksos 
period or at least the beginning of the eighteenth dynasty, that had 
been found along with the Rhind mathematical papyrus. Of a hght 
buff colour and extremely britde, this tightly rolled document had 
been kept in a tin tube since the time of its discovery, as it had been 
found impossible to relax it for unwrapping. Small pieces had 
fractured from the edge of the roU, and these could not be softened 
with water. T his was very surprising, particularly as there was no 
doubt about the identification: it was goatskin, and, on clearing a 
specimen and examining it under the microscope, hairs were still 
visible. Experiments with the fragments showed that, of the film- 
forming media, nitrocellulose was pre-eminent in toughening the 
fibres, and it was decided to apply this to the roll. Also, nitrocellulose 
contracts considerably on drying, and it was hoped that this would 
assist in the unwrapping. 

A syrupy solution of celluloid cuttings was accordingly prepared 
in a solvent consisting of equal volumes of amyl acetate and acetone 
and this was apphed generously to the back of the roU so that it 



soaked well into the skin, leaving a good excess on the outside. While 
plans were actually still being made to control the imwrapping the 
roll proceeded to unwrap itself, the external convolution spreading 
in a wide arc as the celluloid dried and contracted. This movement 
was encouraged by a further apphcation of nitrocellulose to the un- 
treated tissue that had just been exposed. In the course of a couple of 
hours, the manuscript was flat between plate glasses, the celluloid having 
been removed after it had done its work, by washing freely with piure 
solvent. Pis. 6 A and 6 b show the roll before and after unwrapping. 

Professor GlanviUe’s study of this document disclosed that it is a 
hst in duphcate of twenty-six sums in the addition of fractions and 
in one place at least there is an error m addition. It is supposed that 
this scroll had been used for teaching purposes and that scholais 
copied its contents on clay or stone tablets. * 

The second example is that of the Bibhcal fragments of the first 
century b.c. that were found in the cave at ’Ain Feshkha, Palestine, 
in 1949 , accompanying the famous manuscripts known as the Dead 
Sea scrolls. These fragments were of various materials, leather and 
parchment, and were in the form of wads of several layers of the 
membranes, stuck together with the black pitch-Hke material pro- 
duced by the decomposition of some of the leather. The problem 
was to get rid of this black pitch-like material and so separate the 
membranes. It was found possible to relax the fragments by moisture, 
but when damp, the black material became viscous and threatened 
to smear over and obhterate the writing on the more stable frag- 
ments. In this emergency it was found that refrigeration could con- 
trol the stickiness, and while the freezing prevented the tarry matter 
from engulfing the manuscript, it did not immediately harden the 
relaxed tissue. It was thus possible, after relaxing and refrigerating, 
to separate the membranes by the use of a scalpel. After the separa- 
tion had been accomphshed, the membranes were frozen once more, 
and in this state it was possible to cut away and discard the black 
residue which was of no value. 

’ ‘The Mathematical Leather Roll in the British Museum’, Glanville, S. R. K., J. Egypt. 
Arch. 1927, 13, 232. 



All of the materials that have heen worked hy man have at some 
time or other heen used for writing upon — metals, stones, clay, wood, 
ivory, leather — and yet historj^ has been recorded for us, primarily, 
on three that might be considered to be the most ephemeral — papvrus, 
parchment, and paper. These three are completely destroyed by fire 
and water, easily stained, and subject to attack by moulds and insects, 
and yet, when kept dry, they have shown remarkable powers of 

While differing in chemical composition and physical properties, 
they have this in common, that they are brittle when desiccated, but 
when moistened they regain their flexibihty to such an extent that 
they may be safely handled, flattened, and mounted in permanent 
form. The presence of carbon ink does not compHcate treatment, 
and even the more fugitive iron inks, when decomposed in the 
tissue, leave a tracing of rust that is unaffected by the shght degree 
of moisture required for relaxation and manipulation. Parchment is 
a protein substance; papyrus and paper are composed of cellulose, but 
the techniques of conservation, though differing in the three materials, 
are sufficiently related to justify their inclusion together within the 
present chapter. 


The ‘paper reed’ of Egypt, Cyperus papyrus (Lmn.), was grown in 
the delta of the Nile and exploited from earhest times to make ropes, 
sails, boats, matting, cloth, and later to make the paper-hke material 
used for writing purposes. Lucas* describes how he succeeded in 

* Lucas, A., op. cit., p. 164. 



making papyrus sheets. The process he used is as follows. Longitudinal 
strips of the pith of the papyrus plant are laid parallel and shghtly 
overlapping, a second series of strips is placed at right angles over the 
first, and welded to it by beating with a Hght mallet, the only cement- 
ing material being the natural sap, which, in the fiesh reed, is suffi- 
ciently adhesive to effect a permanent join; it acts at the same time 
as a sizing material and prevents ink from spreading when writing 
on the sheet. Pliny refers to flour paste being used to join such sheets 
together to form a long roU. 

Papyrus was used for writings in Egypt from about 3000 B.c. until 
about the ninth century a.d., when the advent of paper-making 
supplanted the manufacture of papyrus as an essential industry. While 
the principal source of documents written on papyrus has been the 
Egyptian tombs, it is known that a form of papyrus was also culti- 
vated in Sicily, and fragmentary remains of charred papyrus rolls 
have been recovered from excavations at Herculaneum. In Egypt 
papyrus writings are not only excavated from tombs, but sometimes 
taken from the mummies themselves. Sir Flinders Petrie found 
old papyrus manuscripts used as layers in the construction of the 
cartonnage of certain mummies of the late Ptolemaic period at 
Gurob. Some of these writings have been salvaged, but from the 
nature of the material and its contamination with plaster, it is not 
surprising that, as documents, they are incomplete and largely in- 

Papyri can always be relied upon to arouse the interest of col- 
lectors, especially when they appear for sale in the form of tight roUs, 
tied and sealed, it may be, with the clay impression of a cartouche, 
perhaps of an eighteenth-dynasty king. It is not unusual to find, how- 
ever, on opening such rolls, that the more attractive the parcel the 
less the interest of the contents. When the roU does contain fragments 
of documents, and this is by no means always the case, they may be 
merely the oft-repeated texts of the Book of the Dead, but where they 
are of hterary texts interest is revived, as Egyptologists have at times 
been successful in discovering the parent document to which the 
missing fragments could be restored. 



Method of unti>rapping papyrus rolls 

Papyrus is recovered from the tombs of Egypt in the form of 
tight rolls, the diameter being determined by the length of the docu- 
ment, and the width varying from a few inches to a foot or more in 
the larger specimens. Such rolls are often parcelled up in strips of 
linen. The hnen may easily be removed, as the flexibUity of the textile 
is unaffected by age, but the papyrus is always very brittle in the dry 
state, and must be relaxed by moisture before any attempt is made to 
unroll it, otherwise it would break into many pieces. The relaxing is 
done by wrapping the roU loosely in several layers of damp white 
blotting-paper and setting it aside for an hour or so on a sheet of 
glass. By this time the external convolutions will have become suffi- 
ciently Hmp to be manipulated without cracking, and imroUing can 
then be commenced. 

Pi. 7 A shows a number of papyri in the dry brittle condition as 
received for unwrapping, and Pi. 7 b shows a stage in the course of 
the work. The damp blotting-paper has been removed from the 
relaxed papyrus and is under the bell-jar to the operator s right; the 
papyrus placed against a sheet of dry blotting-paper is lying on plate 
glass, and is gradually being unrolled, a Httle at a time against the 
paper. When a short length has been exposed, this is covered with 
dry blotting-paper, and flattened rnider glass. After a few inches of 
the roll have been unwrapped the operation is stopped and the im- 
rolled portion covered with the wet blotting-paper to make a fresh 
piece flexible. This is repeated as often as may be necessary, until the 
whole of the papyrus has been unrolled and laid flat. 

Once the papyrus is flat it must be dried without delay, by chang- 
ing the blotting-paper several times in the course of a few hours. As 
a protection against moulds it is usual to sterihze the document by 
pressing it for several days in contact with thymol-impregnated 
blotting-paper (p. 57), after which it may be mounted passe-partout 
between two sheets of glass. 

Comphcations arise when a papyrus is already broken or where it 
is incomplete, and it may then be necessary to get the fragments into 



register by studying peculiarities of grain, colour, thickness, and, of 
course, the writing. Cracks are reinforced with narrow strips of gold- 
beater’s skin; for major repair work either flour paste or a good 
photographic mountant can be used — ^but it is more usual to fix 
detached fragments in position with gold-beater’s skin. Smaller 
papyri may be enclosed in sflk gauze and bound up in a guard-book, 
such as is supplied for newspaper cuttings; this is convenient for study 
as well as for storage purposes. 

Birch bark provides the same problems as papyrus. It is as brittle 
as tinder in the dry condition but it, also, can be relaxed by moisture. 
Pi. 8 illustrates a guard-book with a birch-bark document before 
treatment, and a similar document after flattening and mounting. 
A tightly wrapped wad of writings on birch bark (Pi. 9) was found 
inside a Buddhist image of the eighteenth century. It was possible 
to unwrap it, and to recover eleven legible documents from the wad. 
These documents are copies of sets of Northern Buddhist charms in 
doggerel Sanskrit, adorations of the Buddha, requests for purifica- 
tion, success, &c., in a series of imperatives fortified by magic 


Parchment originated in Pergamon, Asia Minor, in the pre- 
Christian era. It is a stronger material than either papyrus or paper, 
and is able to withstand much harder usage, as one would expect, 
since it is derived from animal skin. Although parchment may be 
prepared from the skin of many animals, the commonest source is 
sheepskin, and this has provided the finest grades for the scribe and 
illuminator. It is, however, almost impossible by examining an old 
parchment to determine with any degree of certainty from which 
animal it has been derived. Exceptionally fine quahties are sometimes 
attributed to the deer, antelope, or smaller animals. In general, the 
younger the skin, the thinner the parchment made from it, and 
the less likely is it to be blemished. Calf parchment is known by the 
manufacturers today as vellum. It is usually harder than sheep parch- 
ment, and on this account of more interest to the binder than to the 


scribe, but the word vellum has lost its definitive character, and one 
often hears the finer grades of medieval parchment described as 
‘uterine vellum’, a description which can seldom be justified. 

Manufacture and qualities of parchment 

The manufacture of parchment involves operations depending for 
their success on the experience and dexterity of the craftsman. Occa- 
sionally one comes across a tour deforce, an old hturgical manuscript, 
perhaps, which in fineness and quahty is a match for the best that can 
be made today, even with the modem abrasive and skin-sphtting 
processes. Fineness is not the only criterion of a good parchment — 
flexibility, perfection of surface, and strength all contribute to 
the excellence of the material as a ground for fine writing and 

The technique of parchment-making has changed little through 
the ages, except for the introduction of the sphtting process. Parch- 
ment was formerly made from the entire skin; in modem practice 
parchment is made from the flesh side of a spht skin. The whole skin 
is limed, dehaired, and very thoroughly defleshed, and a splitting- 
machine separates the skiver from the flesh side. The parchment spht 
is then stretched on a frame and scraped down on both sides with a 
crescent-shaped knife. As it dries the skin tightens considerably, and 
scraping is continued wtith a knife of the same shape but having a 
blade with a burred edge. The parchment is then treated with hot 
water, scraped again, and, while still wet, rubbed on both sides with 
a pumice block. It is then dried on the stretcher frame. 

The type of finish varies in accordance with requirements. In the 
case of writing material which must be white and flawless, bleaching 
and a further degreasing by plastering with hme-wash may be neces- 
sary. On the other hand illuminating parchment must be opaque, 
and the surface is sometimes finished by coating with talc. One of 
the chief aitris of the manufacturer is to suppress the grain, but there 
is usually no difliculty in deciding by inspection which side of the 
material was originally the flesh side, and which the grain side; these 
correspond respectively to the rough side and smooth side of the 



sheet. In good-quality parchment the smooth or grain side is usually 
velvety to the touch, and as it provides a very agreeable surface for 
writing upon, it is preferred for this purpose to the flesh side. It is 
dehcate, in the sense that it is easily damaged by erasures or staining, 
whereas the rough side is more robust and tends in the coarser 
quahties of parchment to be yellowish and homy in character. In a 
bound manuscript both sides are used for writing and, in this case, 
like sides face each other, grain to grain, and flesh side to flesh side, 
so that when the book is open, there is less chance of disparity in the 
appearance of adjoining sheets. 

Parchment may have a transparent quahty resulting firom less 
intensive stretching, or transparency may be obtained by chemical 
means. In 1790 Edwards of Hahfax took out a patent for making 
parchment transparent by treating the flesh side with potash. In the 
King’s Library at the British Museum there is a sample of Edwards’s 
‘transparent vellum’, as it was called, which takes the form of a bind- 
ing for a ‘Book of Common Prayer’. This transparent vellum is 
painted on the inner side and the colours show through the trans- 
parent membrane to the front. It has kept its appearance surprisingly 
well, showing no sign of cockling although it has been on exhibition 
for many years. 

Alkalinity of parchment 

A feature of parchment of all types is that it is alkaline. Liming of 
the skins is carried out in the early stages of feUmongering, prior to 
degreasing, and no subsequent acid treatment is apphed to neutrahze 
the alkah. A trace of hme remains, therefore, and this is firmly held 
by the collagen fibres of which the parchment is composed. This 
alkalinity confers a certain measure of protection against the action of 
moulds and micro-organisms, which prefer a substratum of shghtly 
acid character. It may also provide a clue to the chemical stabflity of 
the material, as parchment is not affected by acidic atmospheres, and 
in this sense is a much more durable material than leather. One dis- 
advantage of alkalinity, however, is the tendency to yellowing, so 
noticeable when parchment is much handled or exposed to dirt and 


grease. The main cause is probably the widespread distribution of iron 
as an ingredient of dust, yielding the coloured hydroxide. This stain- 
ing action is intensihed by damp, and protection from excessive 
humidity is of great importance in the preservation of parchment. 

Moisture sensitivity 

Parchment is a hygroscopic substance and it will adsorb moisture 
in any amount. As an extreme example, if parchment is exposed to 
excess of water for a lengthy period of time, a complete breakdown 
of structure will take place by the chemical action kno\vn as hydro- 
lysis; the proteins are degraded, the organized structures disappear, 
and a form of gelatine results known as parchment size. 

Under normal conditions parchment tends to adsorb and give up 
moisture in sympathy with the rise and fall in the relative humidity 
of the atmosphere. A test appUed to a representative sample of parch- 
ment showed that it contained 10 per cent, of its weight of water 
when in equihbrium with an atmosphere at 40 per cent. R.H.; when, 
without changing the temperature, the atmosphere was suddenly 
raised to 80 per cent. R.H., the parchment adsorbed moisture until, 
in the course of three days, it attained equilibrium with its new sur- 
roundings and the water content was then found to amount to 
no less than 30 per cent, of its dry weight. This test shows how 
responsive the material is to changes in relative humidity, it also 
indicates that there is a time lag. If parchment is exposed to over-dry 
or moist conditions for a short period no harm wiU result, but if 
extreme conditions persist for a time, then deterioration will foUow. 

When kept in an atmosphere that is too dry, say at 40 per cent. 
R.H. or less, parchment tends to become rigid. This condition is 
reversible, however, as flexibility can always be restored by exposing 
the membrane to moisture, but, meantime, damage may have been 
caused to inks and colours through desiccation. Certain inks tend to 
flake from the surface of over-dry parchment and illuminations on 
gesso are hable to suffer in the same way. On the other hand, the 
damage resulting from exposure to high humidity is even more 
severe, especially if the manuscripts are illuminated or decorated with 



painted miniatures. The adsorption of moisture results in cockling 
of the membranes with consequent deformation which may cause 
loss of paint. Under humid conditions miniatures in colour may 
actually become off-set against adjacent leaves. When a bound manu- 
script is exposed for some time in damp surroundings (e.g. in a safe 
or deed box in a damp basement), the edges of the pages pick up 
moisture and become cockled. This type of damage is accelerated if 
the binding is too tight, or if a book has lost its clasp. The book tends 
to gape along the fore-edge, exposing a larger surface area to damp so 
that the wrinkling spreads until, in the end, it is no longer possible 
to close the book. Such distortion cannot be cured by drying or by 
putting the book in a press. The only satisfactory mode of treatment is 
to ‘puU the book’ — a task requiring special training and experience — 
and to deal patiently with the sheets one at a time (p. 85). 

Although so sensitive to moisture changes, parchment retains its 
strength and resihency and is very durable, provided it is not exposed 
to extreme conditions for long periods. The great codices that have 
come down to us from the early centuries of the Christian era are 
often as fresh in appearance as if they had been executed today, and 
although they may have been cleaned, flattened, and rebound, this in 
itself is a testimony to the stabflity of the material. 

Parchment documents that are in a hard and shrunken condition, 
and with individual pages adhering together, are sometimes brought 
in for treatment. In such cases no attempt must be made to separate 
the membranes until they have been relaxed. The relaxing process is 
very similar to that adopted in the case of desiccated papyri (p. 43), 
except that humidification should be less intense, and it should be 
given more time to take effect. The homy tissue of parchment takes 
about twenty times longer to relax than papyrus, and premature 
efforts to separate the leaves would certainly cause damage. In some 
cases it may prove to be impossible to restore the sheets to their 
original size. Even when the membranes are badly shrunken the 
writing may still be quite readable, and it is better to leave well 
alone rather than risk destroying the legibflity of the writing. 

In order to clean a stained vellum binding it is rubbed over quickly 



with a sponge made damp with size water, and if dried at once no 
harm will result. 

The cleaning and repair of manuscripts is a craft that can only be 
mastered by experience. The simple operations of relaxing and 
removing creases, repairing holes, replacing missing comers, &c., 
should be practised on unimportant material. 

For repairs see pp. 85 seq. 

Fungoid attack 

If parchment has become mouldy, the most convenient and effective 
method of fumigation is to expose it in a thymol chamber (p. 56). 
Larger documents may be opened out and reparceUed in contact 
with paper of the same dimensions that has been previously im- 
pregnated with a suitable fungicide (p. 57). This contact method is 
recommended where continuous protection is required over a period. 

Storing parchment 

The control of humidity is an essential factor in the storage of 
parchment. Ideally, the atmospheric humidity should be kept con- 
stant at about 55/60 per cent. R.H. at a temperature between 60° and 
75° F. This will prevent either embrittlement resulting from condi- 
tions that are too dry, or distortion from exposure to conditions that 
are too moist. Damp heat must be avoided as this causes shrinkage 
and encourages the growth of moulds. 

It is no easy matter to maintain satisfactory conditions day and 
night in a private house, or in a safe or strong room, and it would 
have been impossible to suggest a simple remedy to help in this 
matter, were it not for the fact that humidity conditions can be 
stabilized by the presence of a bulk supply of hygroscopic material, 
such as cotton-wool, carpets, curtains, and textiles generally. Shelving 
in an unheated room without ventilation is no place for parchment; 
in the case of metal shelving there will be the added danger from 
condensed moisture. Metal safes and deed boxes are necessary in the 
interests of security, but are not the best accommodation unless plenty 
of hygroscopic material surrounds the parchment to minimize the 

B 6157 




effect of changes in humidity. Documents should be wrapped in 
paper, or preferably in textile, and textiles should be chosen with a 
thought to the subsidiary risk of insect attack. It has been proved that 
cotton and linen materials are safest. 

In storing parchment it should be remembered that parchment and 
papers are prone to attack by rats and mice, and precautions must be 
taken accordingly. Small rodents can work havoc •with parchment 
in a very short time, and to leave a storage drawer open, even for one 
night, may result in damage that is irreparable. Hygiene is the best 
safeguard, and no foodstuffs should be left about that might attract 
the pests. If workmen are on the premises they should be asked to 
exercise care in this respect. 

As to fire precautions, there should always be at hand fire-fighting 
equipment that would not, itself, harm the documents. Carbon 
tetrachloride equipment (such as is used for electrical fires), extin- 
guishers of the methyl bromide type, and extinguishers that emit a 
stream of carbon dioxide gas under pressure are all perfectly safe, 
hut the acid type of fire extinguisher is to be avoided. Asbestos 
blankets should be kept "with the fire-fighting appHances. 


The Chinese claim the honour of being the first to discover how to 
make paper. They teased out silk or vegetable fibres under water, 
and collected the aqueous suspension on a porous support such as a 
stretched cloth. As the water drained away, the support was shaken, 
and this caused the loosened fibres to interlock and so form a thin 
matted sheet. When dry, such a sheet proved to be remarkably 
tough and serviceable. 

The invention is recorded as ha'ving been made at the beginning of 
the second century a.d., and the earhest examples of paper that have 
come do'wn to us date to within fifty years of this time. These were 
discovered by Sir Aurel Stein in the Great Wall of China. Some of 
the actual papers have been examined by the author and were found 
to have been made from rags. This is an interesting fact because the 
fibres of the paper mulberr)’ {Broussonetia papyrifera) were later 


preferred in the Orient, as the raw material for paper-making, and it 
was in the West that paper made from linen and cotton rags was most 
in favour. Hunter, i who has contributed so much to our knowledge 
of the history of paper-making, points out that in China the soft 
absorbent papers made ftom the mulberry fibre commended them- 
selves for brushwork and wood-block printing on one side of the 
paper only, whereas in Europe the requirement was for a hard paper, 
made from rags and treated with size, that could be used for quiU- 
pen work on both sides of the sheet. He remarks upon the fact that it 
took 1,000 years for knowledge of the invention to reach Europe, 
and its progress, via the Persian Trade Route, can be followed by 
studying his dates for the beginning of paper-making in each of the 
countries concerned: 

Year 105 










14th cent. 

France and Germany 









loth cent. 

Damascus and Cairo 



Quality and stability of paper 

The quahty and stabdity of paper is governed in the first instance 
by the raw materials from which it is made, and in Western countries 
many different types of cellulose fibres have been used.^ The strongest 
and most durable papers are hand-made from a mixture of disinte- 
grated linen and cotton rags, the fibres being sized with gelatine, 
whereas the poorest and least durable grades of paper are those made 
by machinery from ground wood, pulped and sized with rosin and 
aluminium resinate. Between these extremes he many varieties of 
modem paper. One of the commonest is composed, for the most 
part, of sulphite pulp, i.e. wood pulp chemically treated in order to 

* Hunter, Dard, Paper-making, 2nd ed.. New York, Alfred A. Knopf, 1947. 

^ Kantrowitz, M. S., Spencer, E. W., and Simmons, R. H., Permanence and Durability 
of Paper (containing annotated bibliography of the technical hterature, a.d. 1885-A.D. 
1939, U.S. Govt. Printing Office, Washington, 1940). 



remove lignin and natural resins. This paper is suitable for book 
prin ting and the printing of half-tone blocks. Modem newspapers,' 
which are not in the same class as regards permanence, contain a high 
proportion of cmde ground-wood fibre, and only a small quantity 
of sulphite pulp. It is easy to determine the source of the constituents 
of paper by teasing out a small fragment in water and examining the 
fibres under low magnification. They may then be identified by 
comparing them with pubHshed photo-micrographs such as are to 
be foimd in a standard reference book.^ 

An estimate can be made of the permanence of paper by various 
forms of testing, e.g. by an accelerated ageing test (heating the paper 
for 72 hours at 100° C.) or by testing the tensde strength or by sub- 
mitting it to an endurance test in a special machine, which folds it 
back and forward, and registers the number of folds that it will 
stand before the sheet cracks. The mechanical tests are often apphed 
after the accelerated ageing test, and of these the endurance folding 
test is the most usefiol. By such testing it can be shown that the 
acidity of paper is of great importance in relation to its permanence. 
Neutral papers (pH 7 i o ‘ 3 ) (see Appendix V ) have a high fold-value, 
while papers that are acidic have a low fold-value in proportion to 
their degree of acidity. Where papers are very acid and the pH is as 
low as 4-4*5, they may actually be too brittle after heating at 100° C. 
for testing in the standard form of folding-machine. 

Acidity may be acquired by paper on ageing, or it may be intro- 
duced in the process of manufacture. Alum is used, for example, in 
rosin-sized papers, and this in itself places them in an inferior category, 
irrespective of the nature of the fibres from which they are made, 
because, in solution, alum is an acidic substance. A residue of bleach- 
ing agent left in any paper will cause deterioration, and the greatest 
care has to be taken in the manufacture of paper to dechlorinate the 
pulp after bleaching. Paper shares with vegetable-tanned leather a 
susceptibihty to attack by sulphur dioxide, and both leather and 

‘ Scribner, B. W., Preservation of Newspaper Records, Misc. Pub. 145, Nat. Bur. of 
Standards, Govt. Printing Office, Washington, 1934. 

2 Armitage, F. D., Atlas of Paper-making Fibres, Guildhall Pub. Co., Epsom, England 


paper have been found to survive best in clean atmospheres free from 
the products of combustion. The mechanism of decay is s imil ar in 
each case, and LangweU^ has shown that sulphuric acid is formed in 
paper, just as limes showed it was formed in leather, by the catalytic 
agency of traces of metals, and he has endeavoured to suppress the 
action of these metals by the use of complexing agents, or inhibitors. 
Iron gall inks actually contain sulphuric acid; this causes brittleness 
and sometimes even puncturing of the paper where there is ink 
writing. Puncturing may also be caused by the acid Hberated during 
the growth of certain micro-organisms (Aspergillus). 

Different methods of overcoming acidity and protecting weakened 
papers have been practised for many years by W. J. Barrow (vide 
infra) who has been concerned not only with the preservation of 
paper, but with the closely related problem of preserving ink writing. 

The exposure of paper to sunlight is another cause of deterioration; 
this is well illustrated by the behaviour of newsprint which becomes 
discoloured and brittle even after a few hours’ exposure. The effect 
is, however, much more general in its application, as has been demon- 
strated by Kimberley and Scribner,^ who found that even the better- 
class papers were weakened by exposure to bright light, and that 
weakening could take place even in the absence of any visible 

Humidity, Mildew, and Foxing 

The control of humidity is of the greatest importance in places 
where books and papers are stored, because paper is a hygroscopic 
material — it absorbs moisture from the atmosphere. Excessive mois- 
ture tends to weaken the tissue, 3 and promotes the growth of micro- 

• Langwell, W. H., ‘The Permanence of Paper Records’, Lib. Asscc. Record, 1953, 55, 
pp. 212-15. See also by the same author Permanence of Papers, Tech. Bull, of the British 
Paper and Board Mfrs.’ Assoc. (Inc.), Part I, 1952, 29, p. 21; Part II, 1952, 29, p. 52; 
Part III, 1953, 30 , p. 2. 

^ Kimberly, A. E., and Scribner, B. W., Summary Report of National Bureau of Standards 
Research on Preservation of Records, Misc. Pub, M 154, Govt. Printing Office, Washington, 
D.C., 1937. 

3 Carson, F. T., Effect of Humidity on Physical Properties of Paper, Circular C. 445, 
National Bureau of Standards, Govt. Printing Office, Washington, 1944. 



organisms which rot the size and cause staining. How deHcate the 
hygrometric balance is may be gauged from a calculation that was 
made in connexion with an extension of a quadrant book-stack at 
the British Museum. One thousand tons of books were calculated to 
absorb at least 20,000 lb. of water when the relative humidity of the 
atmosphere increased from 57 per cent, to 63 per cent, at 60° F.^ To 
avoid deterioration, books and papers should be stored under con- 
ditions that are constant, say 60 per cent. R.H. at 60° F. 

Cellulose and gelatine size, as well as the constituents of book- 
binder’s paste (starch, flour, or dextrin), are all excellent nutrient 
materials for fungi, and when the relative humidity of the atmosphere 
is over 70 per cent., the appearance of mould growths on paper is not 
long delayed. If there is an outbreak and the trouble is detected and 
dealt with at once by drying the affected papers httle harm wfll 
ensue, but if infection is allowed to continue unchecked, the papers 
will be yellowed and stained with coloured spots. Some micro- 
organisms hve on the size, and others on the cellulose fibres. Where 
the sizing material has been attacked, the size is destroyed, and the 
affected area becomes highly absorbent, like blotting-paper, and if 
tested by wetting, will appear dull and translucent as compared with 
the surroundings. When cellulytic micro-organisms are present, the 
surface of the paper will be eroded, and the paper may become very 
brittle. In both cases there is a tendency for iron salts (iron is a com- 
mon impurity in paper) to be accumulated on the damaged areas, 
where they form rusty brown spots commonly known as foxing. 
But this type of staining may arise from other causes,^ and is often 
due, in part at least, to the coloured products elaborated by organ- 
isms in the course of their development. 

Methods oj dealing xvith an outbreak of mould growth 

When mould begins to grow, 3 the first indication is the appearance 

' McIntyre, J., Journal oj the Institution of Heating and Ventilating Engineers, 1937, 4, 
No. 48, p. 570. 

^ Armitage, F. D., The Cause of Mildew on Books and Methods of Protection, Bull. 8, 
P.A.T.R.A., Lcatherhead. 

^ Plenderleith, H. J., ‘Mould in the Muniment Room’, Archives, 1952, 7, p. 13. 


of a fine white fluff, not at first easy to discern, but soon forming 
furry patches of roughly circular shape, and if no action is taken, 
growth, once started, will be hkely to proceed apace. The real source 
of the trouble is damp, but so long as infected material is in the room, 
it would be fatal to attempt to deal with the outbreak merely by 
raising the temperature, as this would stimulate growth and spread 
infection. A prior requirement is the improvement of ventilation by 
the use of fans, &c., special attention being paid to any damp spots. 
Moulds must not be brushed off the books in the room; all mouldy 
material should first be removed, and then the room can be dried by 
electric heaters and fans, or by the dual-purpose machine known as a 
hot-air circulator, which is merely a magnified form of the famfliar 
electric hair-drier. 

Infected books and manuscripts can best be dealt with in the open 
air, or in a large airy room, by brushing and sunning, books being 
stood on edge, open, to allow firee access of air, and documents 
flattened and suspended on stretched lines. 

If the damp condition of the room can be permanently cured, and 
if the above instructions have been carried out, all the material may 
safely be returned and no further outbreak should occur, but if there 
is any doubt about dry conditions being maintained, then sterfliza- 
tion will be necessary. 

(A) Fumigation of room. Gaseous formaldehyde may be used for 
sterilizing a room containing mould-infested papers,* provided the 
room can be effectively sealed for this purpose. The gas is generated 
by adding aqueous formaldehyde to a substance such as potassium 
permanganate, with which it reacts exothermically. Details of the 
procedure are given by Walker, ^ who recommends adding i lb. of 
aqueous formaldehyde (Formahn) to about 6 oimces of soHd per- 
manganate in a large porcelain basin or bucket. This quantity is 

’ Protein materials (parchment, vellum, leather) should, preferably, not be present as 
these tend to become hardened by the action of the fumigant. Despite this warning there 
may be cases in which the fumigation in situ by formaldehyde vapour could be justified 
as this is the most direct way of preventing moulds from spreading. 

^ Walker,}. F., Formaldehyde, American Chemical Society Monograph, 1944, p. 327. 



sufficient for a room of 1,000-1,500 cub. ft. As the gas is evolved 
briskly, the operator should make a quick exit, and seal the door on 
the outside. After twenty-four hours or longer, the room should be 
thoroughly ventilated, and any lingering odours of the fumigant 
dispersed by sp rinklin g the floor with ammonia, which converts the 
residual formaldehyde to hexamethylenetetramine, which is odour- 
less. Alternatively, the sterilization may be carried out using a para- 
formaldehyde heater, 5 oz. of commercial paraformaldehyde being 
required for a room of 1,000 cub. ft. 

(B) Sterilization of materials. Two methods are available for steriliza- 
tion — flunigation which is effective but confers no lasting protection, 
and a method involving the use of papers impregnated with fungicide 
of low vapour pressure which gives continuous protection over 
a period. 

(i) Fumigation with thymol vapour. A suitable chamber for this pur- 
pose may be improvised from any relatively air-tight cupboard, and 
to be generally useful it should be large enough to accommodate in a 
horizontal position anything up to an imperial sheet (30 by 22 in.). 
The material to be sterihzed is supported on a framework (stretcher 
covered with a net or strands of tape or twine), some 2 feet above 
the bottom of the cupboard. A 40-watt electric lamp is fixed near 
the bottom of the cupboard, and this emits enough heat to melt the 
thymol crystals, which are placed some 2 inches above it, in a clock- 
glass or enamel plate, supported on a wire stand. About i ounce of 
thymol is required for sterflizing the contents of a cupboard of 
16 cub. ft. capacity. 

To use the apparatus, the current is switched on for two hours, 
then switched off, and the heating should be carried on for periods of 
about two hours every morning for fourteen days. If much material 
is being dealt with, the papers should be rearranged in the cupboard 
each morning before the hght is switched on. After each treatment 
the door of the cupboard should be kept closed for about twenty- 
four hours. 

In stacking the cupboard it is important to arrange for free access of 
the thymol vapour around the infected material — small manuscripts 


may be suspended, rolls set up on edge, books stood on end with their 
pages open fan-wise, &c. The success of the treatment depends on 
two factors, the concentration of thymol vapour reaching the fungus, 
and the drying action which attends the sHght rise in temperature 
from the electric-Hght bulb. 

It has been found by experiment that the method may be safely 
appHed to prints, drawings, manuscripts, pastels, water-colour paint- 
ings, books, and also to parchment and vellum. No harm has been 
found to result from prolonged dosage. It should be noted that od- 
paints and varnishes become softened under treatment, and for this 
reason the method is unsuitable for the sterilization of easel paintings 
and the hke. The thymol chamber itself should of course be free from 
any internal paintwork. 

(ii) Fumigation with formaldehyde vapour. Formalin (40 per cent, 
formaldehyde)^ has long been known as a powerful antiseptic and 
germicide. It may be employed for sterilizing papers, but is not 
recommended for parchment, vellum, or other protein products, as 
these are hardened by this reagent. The papers are exposed for about 
twelve hours over a dish containing formahn (about 2 oz.). The 
sterilization must be carried out in a sealed air-tight box, and during 
exposure the temperature inside the box should not be allowed to 
fall below 65° F. It is important also that the relative humidity should 
be kept above 60 per cent, as moisture is essential for effective dis- 
infection. After sterihzation the papers should be freely exposed to 
the air for several hours. 

(iii) Use of impregnated paper. White blotting-paper may be im- 
pregnated by immersing it for a moment in a 10 per cent, solution of 
thymol in alcohol, after which the excess of solvent is allowed to 
evaporate, leaving the thymol uniformly dispersed in the sheet. 
Sheets with a higher but less uniform concentration of thymol are 
made by scattering a handful of thymol crystals between several 
layers of absorbent paper and melting the crystals into the tissue by 
the appHcation of a hot electric iron. Thymohzed papers, prepared by 

* As formalin vapours are highly irritant, precautions must be taken not to inhale 



one or other process, are used for sterilizing papyrus after it has been 
unroUed, or for interleaving mildewed books. 

For the permanent protection of large parchment rolls in store, a 
fungicide should be chosen that is less volatile than thymol and more 
potent in its action. In this case the quantity of fungicide can be 
greatly reduced, and the paper used for impregnation may be thin 
and not greatly absorbent. Protective sheets are prepared by passing 
a roU of thin paper through a 10 per cent, aqueous solution of the 
sodium salt of pentachlorophenol (Santobrite). Such paper provides 
a valuable means of protecting hbrary material and suchlike in 
tropical countries where the incidence of mould growth is a serious 
factor in conservation. 

Insect pests 

In common with other organic materials, parchment and paper are 
attacked by certain insect pests, and there is no doubt that the pre- 
sence of gelatine and starch increases the number of potential enemies, 
and that the greatest activity is associated with damp. 

McKenny Hughes* has described the insect pests of books and 
papers common in Britain in a communication to the technical sec- 
tion of the British Records Association. He lays emphasis on clean- 
liness and free circulation of air as being the best deterrents, and 
suggests that infected manuscripts can be most simply and effectively 
sterihzed by exposure in an air-tight cupboard containing crystals of 
paradichlorobenzene. The crystals should be present in a quantity of 
I lb. to every 10 cub. ft. of air space; the cupboard should be sealed 
with gummed paper, and the time of exposure should be for at least 
a fortnight at a temperature of not less than 70° F. 

When attack by insects is a recurrent problem, the vacuum chamber 
method of fumigation has much to recommend it. This has been 
apphed on an extensive scale for the sterilization of rare books and 
manuscripts in the Huntington Library, San Marino, Cahfomia,^ 

* McKenny Hughes, A. W., ‘Insect Pests of Books and Paper’, Archives, 1952, 7, p. 19; 
see also Weiss, H. B., and Carruthers, R. H., Insect Enemies of Books, New York Public 
Library, 1937, which contains an annotated bibhography of the subject to 1935. 

^ liams, T. M., Library Quarterly, 1932, ii, 4, p. 375. 


using ethylene oxide gas, which has been foimd to be a safe and 
effective fumigant. 

When an attack by insects has been discovered, the accommoda- 
tion — drawers, cupboard, shelving — should be thoroughly cleaned 
and dusted with some persistent insecticide such as DDT, or BHC 
(Gammexane). Chakravorti^ has warned against exposing the papers 
themselves to Gammexane smoke, as this brings about a substantial 
decrease in tensile strength and folding endurance, and tends to cause 
iron gall ink to fade. 

Resizing and bleaching 

When paper has been weakened by the decay of the sizing mater- 
ial, it loses its characteristic ‘ratde’, becoming hmp and absorbent, and 
in this condition is readily stained. It may be resized (p. 87), but it is 
often desirable to clean it beforehand by a process of bleaching 
(pp. 75 seq.). 

Bleaching and pressing are operations carried out in the normal 
course of manufacturing paper, and they are among the commonest 
procedures in restoring old papers. Solutions of hypochlorite were 
at one time employed, exclusively, for bleaching pulp in the manu- 
facture of paper, and also for cleaning stained prints in the studio. In 
1937, chloramine-T^ was advocated as a less drastic agent for use in 
restoration work (p. 77). More recently a method was described by 
Gettens^ making use of sodium chlorite, and while this eliminates the 
risk of oxidation and the consequent weakening of the tissue which 
is always present when hypochlorites are employed, it requires the 
use of apparatus which is not normally available in a restorer’s labora- 
tory. It marks such an important advance in the technique of bleach- 
ing, however, as to merit every prominence. 

Gettens gives details of three methods of apphcation suitable for 
the treatment of various types of material, and these have all been 

' Chakravorti, S. K., Nature, 194.9, PP- 607-8. 

* Plenderleith, H. Conservation of Prints, Drawings and Manuscripts, 1937. 

^ Gettens, R. J., ‘The Bleaching of Stained and Discoloured Pictures on Paper with 
Sodium Chlorite and Chlorine Dioxide’, Museum, 1952, 5, No. 2, p. 116. 



tested in the British Museum Laboratory and found to be satisfactory. 
It will suffice to describe the simplest method which is recommended 
for bleaching prints, engravings, etchings, and drawings in carbon 
ink or in pencil, when these may safely be immersed in water. This 
method can be applied, provided an efficient fume cupboard is avail- 
able in which all operations can be conducted. It is important that the 
cupboard be well ht so that the progress of the reaction can be ob- 
served through glass, and it should be provided with a drain and 
running water as all operations must be carried out in the closed 

To prepare the bleaching solution 75 ml. of 40 per cent, formalde- 
hyde (Formalin) are added to a 2 per cent, aqueous solution of sodium 
chlorite (made by dissolving 60 grams of technical sodium chlorite 
in 3 htres of water) in an enamel photographic tray of suitable size. 
The solution is observed to become yellow owing to the formation 
of chlorine dioxide which is the active bleaching-agent. The stained 
print is laid on a glass plate to act as a carrier, and the whole im- 
mersed in the solution until the stains are removed. This will take at 
least fifteen minutes, but may run to upwards of an hour, depending 
on the nature of the stain. The concentration of the bleaching solu- 
tion is not critical; it may be weakened or strengthened and 10 ml. 
of a wetting-agent such as Lissapol added if desired. 

When bleaching is completed, the print, resting on the glass plate, 
is removed from the chlorite solution and washed in running water 
for at least fifteen minutes to remove sodium salts, no intermediate 
antichlor bath being required. The print should not be lifted from 
the glass support while stiU wet, otherwise it is Hable to be damaged. 

It has been found that this process removes water-stains, fox-marks, 
and mildew, and does not give the paper a staring white appearance, 
as is the tendency with processes involving the use of hypochlorite. 

Strengthening and Deacidification 

When papers are damaged or very brittle they may be strength- 
ened by the process known as lamination. Lamination may be carried 
out either by covering both sides of the paper with fine silk of open 


weave (Crepeline), using a starch paste or dextrin adhesive, or by 
sandwiching the paper between sheets of cellulose acetate, using a hot 
press, without the apphcation of any adhesive. Laminated papers of 
various kinds have been tested by Scribner, ^ who found that the 
sheets were strengthened without impairing the stabiUty of the 

The silk method, which requires no special equipment, has been 
much employed for the repair of valuable books and manuscripts, 
but it may add considerably to their weight and bulk, and the 
adhesive has an obscuring effect, limiting subsequent study by photo- 
graphy. The paste fluoresces when the paper is examined by ultra- 
violet radiation, and this is a great disadvantage. The sflk, moreover, 
is d iffi cult to remove when, in time, it requires to be replaced, and as 
this may be, on a conservative estimate, every twenty-five years, 
there are some serious objections to ‘silking’. By contrast, the plastic- 
sheet method has the great advantage of speed in processing, the 
cellulose acetate film is transparent to all radiations, and it may be 
readily removed at any time without straining the paper merely by 
passing it through baths of acetone. There need be httle change in 
the appearance of the paper when laminated according to the cellu- 
lose acetate method, and there is the minimum of increase in the 
weight and bulk. 

Any hesitation in recommending the cellulose acetate process for 
the lamination of valuable documents is concerned with the fact that, 
while cellulose acetate may be accepted in itself as innocuous and 
durable, the sheets used for laminating contain a relatively high 
percentage of plasticizer, and it is by no means certain that this will 
remain in the film indefinitely, or indeed that one or other of the 
plasticizers used may not in some as yet undefined way be deleterious 
to the paper. It is only fair to add that in the Barrow^ process which 

' Scribner, B. W., Protection of Documents with Cellulose Acetate Sheeting. Govt. Printing 
Office, Washington, 1941. 

^ Barrow, W. J., ‘The Barrow Method of Laminating Documents’, J. Documentary 
Reproductions, 1939, ii, 147; ‘Restoration Methods’, The American Archivist, 1943, vi, 151; 
Procedures and Equipment used in the Barrow Method of Restoring Manuscripts and 
Documents, W. J. Barrow, State Library Bldg., Richmond, Virginia, U.S.A., 1950. 



employs no auxiliary adhesive, and which has been tested on a large 
scale since 1939, there is no evidence that these criticisms have any 
foundation. It would nevertheless be reassuring if a standard speci- 
fication were forthcoming for cellulose acetate film suitable for the 
lamination of valuable documents. 

There are two stages in the Barrow process, (i) removal of the 
acidity of the paper, and (2) lamination. These two operations are 
complementary and are designed not only to ehminate existing acid- 
ity but to protect the paper firom acid attack in future. The lamina- 
tion process confers mechanical strength and may also have the effect 
of intensifying faded writing. Details are as follows: 

1. Removal of acidity. The paper is placed between two smooth 
flexible grids of copper, so that it can be conveniently manipulated, 
and immersed for twenty minutes in a saturated solution of lime 
water (c. 0-15 per cent, strength). The acid in the paper is neutralized 
and the paper then contains a shght excess of lime. It is transferred to 
a bath of calcium bicarbonate (0-20 per cent, strength) for a like 
time. In this bath the excess of hme is converted into calcium 
carbonate (chalk) which is finely precipitated in the fibres of the 
paper where it is retained. The actual quantity of chalk is very 
small, but sufficient in amount to confer protection against any 
further acid attack to which the paper may be subjected in urban 

2. Lamination. After the acidity has been removed and the paper is 
dry, it is placed between two sheets of cellulose acetate (thickness 
0’002 inch, or for dehcate work o*ooi inch), and sheets of fine tissue- 
paper are laid on either side making a five-ply structure. This is 
placed in the Barrow laminator, the papers and tissue preheated 
for a few seconds, then hot-pressed at 315-25° F. by passing through 
a pair of calender rolls from which a single laminated sheet emerges. 
The tissue-papers seem to disappear in the process of lamination as do 
the acetate films, and fine writing is made rather more legible owing 
to the high refractive index of the cellulose acetate with which it is 
now consohdated. Lamination by this process causes Httle change in 
appearance but there is a marked improvement in durabflity. 


The Barrow laminator is a precision instrument consisting of a 
pair of thermostatically controlled preheating plates, and a pair of 
calender roUs, between which may be appHed a pressure varying 
from 300 to 2,000 lb. per sq. in. When the papers and films are 
assembled, a sheet may be laminated in about thirty seconds. Inci- 
dentally, the equipment provides a ready means of mounting maps 
on cloth, and if the map is surfaced with cellulose acetate film (but 
not with tissue-paper) it becomes impervious to ink stains and 


Having considered the principal materials used as grounds for 
writing, attention must now be given to inks.' 

The oldest writing-ink of any permanence is carbon ink, which, in 
its primitive form, was no doubt a mixture of fine soot held in sus- 
pension in oil, gum, or glue size. Carbon is fast to Hght, and as it is 
unaffected by chemical agents, it has lasted well. 

The traditional writing-fluid in Ancient Egypt was a carbon ink. 
It was used on papyrus, wood, cartonnages, and potsherds. When 
papyrus is exposed to the sun, the natural biscuit colour of the reed 
is gradually bleached to a pale ivory, and as any writing in carbon ink 
is unaffected, the net result is that the ink appears to be darker and 
altogether more intense against the Hghter ground. The binding 
medium of the ink (gum or glue) has probably long since decayed, 
but the carbon particles are locked permanently in the tissue and 
remain black. Chinese and Indian inks are carbon inks, and no prob- 
lem is involved as regards their preservation, except when they 
happen to have been apphed thickly to a non-absorbent ground, so 
that the ink adheres badly and tends to flake off. In this case the 
pigment may be fixed by applying thin gelatine size (p. 87). Carbon 
is the basis of our modem printing inks, and of the black uiks used 
in the etching and engraving processes. Carbon inks are still preferred 

' For a brief history of writing inks, especially of the iron-gall type, see Waten, C. E., 
Inks, Circular C. 413, National Bureau of Standards, Govt. Printing Office, Washington, 
U.S.A., 1936. 



for permanent writing today, and, as a class, they are referred to as 
manuscript inks. Black waterproof drawing inks are usually carbon 
inks containing a little varnish. 

Carbon inks in general are all unaffected by the mild cleaning and 
bleaching processes carried out in the laboratory. 

Iron inks are in a different category. They are compounded from 
gaUo tannic acid in presence of iron, and as the tannates are obtainable 
firom a variety of natural sources, the inks vary in quahty. Also from 
early times iron inks have been prepared according to a multitude of 
household recipes, and this is another reason for the wide variations 
in substance and permanence. In some cases the black is indistinguish- 
able in appearance from carbon; in other cases the ink may be a 
rusty brown or yellow, and so faded as to make the writing illegible. 
It is not safe to assume that because a sample of writing is ancient and 
dark in colour, it must necessarily be in carbon ink. Iron inks have 
been identified which date to the second century b.c., and it is 
reasonable to suppose that they may have been in use before this; 
their discovery would approximate in time to the discovery of tan- 
ning as a means of preserving animal skins from putrefaction. 

The acidity of iron-gaU ink may be due to tannic acid or sulphuric 
acid which are present in varying amounts. Writings have been ex- 
amined in which the acidity has caused the paper to become perfor- 
ated and reduced to a frail membrane of lace-hke appearance. This 
phenomenon is not confined, however, to iron-gaU inks. Any acid 
ink may destroy the tissue in the same way, and as all sorts of mix- 
tures have been employed as writing-inks, it is not surprising to find 
occasionally that even a carbon ink has been sufficiently acid to 
deteriorate the paper. 

Iron inks are recognizable by a simple chemical test. When a tiny 
drop of I per cent, acetic acid is placed by capillary tube on dried 
iron ink, the ink will become locally soluble; the drop is absorbed 
into filter paper and it will be turned to Prussian blue if treated with 
a drop of i per cent, potassium ferrocyanide. It is seldom, however, 
that such a test is called for. In bleaching a print which bears a 
signature, it is safest to assume that the signature is in iron ink and 

The d^ht roll ot birch bark ( i j) as removed trom the inside ot a Buddhist image 

Eleven thin birch bark sheets, inscribed in doggerel Sanskrit, recovered from the 

roll. (Xi) 


10. 1>A(,I, {)! DltAWINCS liY l.I.ONAIiDC) DA VINC 


therefore fugitive, and to protect it before treatment by a thin wash 
of celluloid (5 per cent.) dissolved in a solvent composed of equal 
volumes of acetone and amyl acetate. The film of nitrocellulose pre- 
vents the action of chemicals on the inks, and may easily be removed 
after treatment when the drawing is quite dry, by an apphcation of 
a httle of the solvent, the area being pressed afterwards with white 

As carbon is insoluble, the ink made by grinding it in a binding 
medium cannot be other than a suspension of fme particles, so that 
when appHed to paper the medium will tend to be absorbed, leaving 
the particles in the interstices near the surface. With iron inks there 
is a greater tendency for the pigment to spread, ' and, indeed, it is the 
flow of such inks that commends them for use with the pen, whereas 
in the case of carbon inks, the brush is the more suitable instrument 
for calhgraphy. 

In studying yellowed inscriptions that seem to be executed in iron 
ink, one should bear in mind that artists and writers have often been 
attracted by rusty colours, and at times have chosen to write or draw 
in materials which are recognized today as being impermanent. Such 
are the inks made from sepia (cuttle-fish ink), and from bistre (beech- 
wood soot). In cleaning and restoration work it is safest to assume 
that aU pale rusty ink lines are fugitive. Coloured inks also are all 
fugitive. The most stable appear to be the reds. These vary in com- 
position depending upon their source: some, such as madder and 
logwood, are extracted from plants, and others are obtained from 
insects, for example, kermes from Persia, grain from Poland, and 
cochineal from Mexico. The so-called Tyrian purple comes from 
several species of shell-fish. But even the reds are evanescent, and 
care should be taken not to expose them to any form of treatment 
that might result in discoloration or loss. 

To try to protect any of these fugitive colours by applying a thin 
wash of nitrocellulose is not recommended, as in some cases celluloid 
solution might cause the colours to run. 

' For the effect of the migration of inks, &c., on the permanence of paper see Barrow, 
W. J., Archivum, 1953, 3, p. 105. 

B 6157 




Reading faded writing 

Chemical methods of intensifying writing should not be used until 
the possibilities of obtaining information by modem photographic 
processes have been exhausted. 

When it is desired to read writing that is illegible, it is first exam- 
ined by different forms of illumination with the help of hght filters, 
in the hope that by this means details will be revealed which are not 
visible in the ordinary way. 

A source of ultra-violet illumination in a darkened room is some- 
times particularly valuable in making indistinct writing readable, and 
at other times the infra-red viewer may give results that are otherwise 
unobtainable. When an inscription is partially legible under ultra- 
violet hght, the chances are that a photograph taken using this source 
win be even more rewarding; and likewise, should the infra-red viewer 
afford any suggestion that the writing is intensified, it will be worth 
while taking photographs on specially sensitized infra-red plates using 
a tungsten lamp source. It is quite impossible to predict in advance 
which method, if any, is likely to be the more effective, and the 
results are often entirely negative. It may be stated, in general, that the 
ultra-violet type of examination is more useful than the infra-red, 
though with dark subjects the latter is sometimes strikingly successful. 

Photographic processes are of great value in studying faded or 
damaged drawings as they provide a means of intensifying faint lines 
and suppressing stains, in this way making it possible to appreciate 
the artist s work in facsimile. Fluorescence photographs obtained by 
using a filtered ultra-violet source occasionally yield surprising re- 
sults as in the case of certain pages from a sketch-book of Leonardo 
da Vinci in the Royal Collection at Windsor Castle, one of which is 
illustrated in Pis. ioa and b. An ordinary photograph taken on a 
panchromatic plate recorded the drawing of a hand. Inspection by 
filtered ultra-violet radiations indicated that several drawings were 
still extant on the same page and it was possible, by photographing 
the page under ultra-violet illumination, to obtain a complete record 
of the drawings. 


Charred documents. When a document has been carbonized by fire, 
it may still be possible to photograph the writing by dayhght using 
a high-contrast blue-sensitive plate^ or by ultra-violet or infra-red 

In some cases the writing may be rendered sufficiently distinct to 
be read by immersing the paper in a 5 per cent, solution of silver 
nitrate for about three hours, when the writing will appear black on 
a grey ground.^ The method which seems to have met with most 
success, however, is that of Taylor and WaUs^ in which the docu- 
ment is given several appfications of a chloral-hydrate solution (25 
per cent, in alcohol), and dried at 60° C. between each apphcation. 
It is then treated with glycerine (10 per cent.), dried as before, and 
photographed, using a contrasty non-colour sensitive p’ate. It is 
claimed that this method, even in the case of typescript or printing, 
has never failed to give a readable result. A comprehensive summary 
of the methods available has been pubhshed by the Istituto di Pato- 
logia del Libro, Rome.'^ 

' Jones, G. A., ‘Deciphennent of Charred Documents’, Nature, 1941, 147, pp. 676-7. 

* Murray, H. D., ‘Examination of Burnt Documents’, Nature, 1941, I48> p- I99- 

^ Taylor, W. D., and Walls, H. J., ‘A New Method for the Decipherment of Charred 
Documents’, Nature, 1941, 147, p. 417. 

* Santucci, L., ‘Metodi per la rigenerazione di documenti carbonizzati’, Bollettino Jell’ 
Istituto di Patologia del Libro, 1953, 12, pp. 95-102. 




In the conservation of prints, drawings, and manuscripts the practical 
problems of cleaning, repair, and mounting are numerous and varied. 
While the more complex operations and dehcate manipulative pro- 
cesses require the speciahst, a number of simple methods remain that 
can be apphed in an improvised laboratory by the amateur, and it is 
the purpose of this chapter to present a series of practical instructions 
for the collector who may be interested to carry out some of his own 

Successful restoration depends on practice and experience. Old 
and worthless prints and drawings should be collected for experi- 
mental purposes, and repeated trials made in order to acquire dex- 
terity in handling dehcate material, and to discover the merits and 
limitations of the various processes described. Reagents should be 
tested on inconspicuous parts of the work before applying them 
generally, the milder processes being chosen initially and the more 
drastic ones only when necessary, and after carefully considering the 
risks involved. Thus weak solutions are tried before strong, cold 
solutions before hot, and iruld reagents before those known to be 
more powerful in their action. Although technical details are given, 
the various processes may require modification according to the type 
of material undergoing treatment, and it is to be understood that the 
methods only apply, unless otherwise stated, to the more stable forms 
of art — prints, engravings, etchings, and drawings in carbon ink or in 
pencil when the design is not affected by immersion. 

A large shallow sink provided with hot and cold running water is 


an essential for print cleaning but, beyond this, no elaborate apparatus 
is required. The smk should have a draining rack on one side, and be 
flanked by a fair-sized table on the other, the table having half its 
surface covered with |-inch plate glass held permanently in position 
by a wooden moulding of the same thickness. It is convenient to 
have heating facihties for paste-making in the same room. The room 
should be well ht with both natural and artificial hghting. Beyond 
storage accommodation for papers and chemicals the following com- 
pletes the hst of requirements — 

Double saucepan and jug. 

Measuring glasses. 

Sponge and glass cloths. 

Large photographic trays (porcelain and enamel). 

A quantity of good-quahty white blotting-paper. 

Drawing-instruments, rules, squares, &c. 

Drawing-boards and drawing inks. 

Architect’s soft erasers. 

Flexible paper-knives. 

Print-trimming knife and oilstone. 

Paired sheets of plate glass of varying weight and size. 

A few heavy weights. 

Brushes for dusting, retouching, and applying paste. 

Electric iron. 

Adjustable reading-lamp. 

Copying-press and a capacious fume cupboard — optional. 


The paper. Examine by transmitted and reflected Hght, using a lens 
when necessary. Test the crackle on shaking. 

{a) Is the paper very porous, soft, or spongy? It may be so soft that 
it would be unwise to use immersion methods of cleaning (see 
Japanese Prints, p. 82). When paper is moistened it expands, the 
effect of any size or binding material is weakened, and, if carelessly 
manipulated in this condition, it may easily be tom. 

{b) Is the paper hard or brittle, or the surface pitted or rotten? The 
hardness due to size must not be mistaken for strength — test in a 



comer with water. Papers which have been in the tropics are fre- 
quently so brittle that they will not stand manipulation without 
fracture. When the surface is pitted, ink may be inclined to flake, 
and immersion methods should be avoided lest the ink or paint 
should float away. Take note of weakened areas, creases, tears, worm- 
holes, &c. 

The drawing or impression. To what category does the print, draw- 
ing, or manuscript belong ? What is the nature of the ink (pp. 63 seq.) ? 
Is the technique simple or compound ? Is any water-colour present ? 
Does the picture bear traces of having been retouched or treated 

The object to be treated must be carefully examined in every 
detail before a course of treatment can be decided upon. 

I. Removing cardboard backing 

The card is always removed from the print: never the print from 
the card. 

Cardboard is composed of laminated sheets, and the first opera- 
tion is to insert a knife at a comer to determine the number of 
layers. A long flat paper-knife may then be pressed in and the 
laminations successively removed from the back, leaving about two 
of the constituent sheets attached to the print. In this condition the 
print is held with the back in the steam of a boiling kettle until the 
card becomes quite soft. It is then laid face down on dry clean 
blotting-paper and the card drawn slowly away by dragging gently 
on a comer across the part as yet undetached. In this way, if the card 
has been sufficiently steamed, it may be removed without str ainin g 
the print. The print is now laid face down on fresh blotting-paper, 
and any residual adhesive carefully removed Avith a clean sponge. If 
this operation is omitted, the paper will eventually cockle. It is then 
allowed to dry between fresh sheets of blotting-paper under a 
weight such as a sheet of plate glass. 

W^hen, as sometimes happens, the backing is an inferior brown 


strawboard, nothing but prolonged washing wiU soften it. This type 
of board is not laminated and can only be removed by rubbing it 
gradually away with the fingers while wet. 

2. Removal of paper backing 

When completely pasted down on paper, a print cannot generally 
be detached by steaming. Lay it face down on glass, sponge the back 
of the mount with warm water. Now float the picture face upwards 
on lukewarm water and allow plenty of time for the adhesive to 
soften before attempting to detach the moimt. When starch paste is 
hard and slow to respond to such treatment, softening may be facih- 
tated by floating the drawing face upwards on a warm aqueous solu- 
tion of an enzyme (Appendix I. 3). Remove residual adhesive and 
complete operations as in i. 

3 . Removal of canvas hacking 

When a print has been pasted down on canvas fixed on a stretcher, 
a sharp knife is inserted at one comer and drawn along the edges to 
remove the canvas and print firom the wooden firame. 

The back of the canvas is then sponged with lukewarm water, and 
laid against a wet sheet of glass, the fabric being made as wet as pos- 
sible and the print kept as dry as possible. After some time when the 
adhesive is st^ciendy soft, clean blotting-paper is laid on the print, 
and the blotting-paper, print, and canvas inverted and placed for 
support on another sheet of glass. The canvas is now worked back 
from a comer by gently pulling it across the part as yet undetached. 
If the paper starts to split, and shows signs of partly coming away on 
the canvas, stop immediately, apply very hot water to the canvas, 
and wait for some time before proceeding. 

Complete the operation as in i above. 

4. Removal of varnish 

A print varnished with an oil varnish may be irretrievable as this 
type of varnish becomes insoluble with age. The removal of spirit 
varnish is possible, but is rather a specialized operation requiring 



considerable care. The treatment given below may have to be modi- 
fied to suit certain types of print. 

Rub a tuft of cotton-wool in the palm of the hand (to give it a 
smooth surface), damp it, and hghtly rub over the varnished surface. 
After drying, repeat the operation using turpentine instead of water. 
This clears the picture, and a record should be made of any colours 
before proceeding further. 

In order to determine the most suitable solvent for removing the 
varnish, test a comer of the print with some methylated spirit. If 
this is not effective, try hquid ammonia (o-88) diluted with water 
(i: 50). Support the print on a glass plate face up, and flood it with 
the chosen solvent, the action of which may be assisted by a flat 
camel-hair brush. Fresh solvent should be added repeatedly until it 
no longer becomes stained with the varnish. The print is now free 
from varnish although stfll more or less tinged faint brownish-yellow 
because the cellulose fibres have been stained with the darkened 
varnish. The print must be rinsed with water, then bleached {vide 

Plate II illustrates how, in the case of a printed book, it was 
possible to remove intensive varnish staining by the solvent action 
of alcohol followed by bleaching 

Old spirit varnish may be very brittle and may resist the above- 
mentioned solvents. However, hot water will sometimes soften it so 
that it flakes away. Pour boding water into a large enamel photo- 
graphic dish and immerse the print. As this method cannot easdy be 
controlled where colours are concerned, it is risky with anything 
other than an engraving. 

cleaning: dry and wet methods 
I. Dry methods 

Although dry cleaning is not always necessary, pencil marks, &c., 
may become fixed in the paper if this is omitted before wet treatment 
is carried out. 

If mddew is present, pick off the fluffy surface growths with a soft 


camel-hair brush, care being taken that the spores are not scattered 
about in the process. Many are unavoidably rubbed into the tissue of 
the paper. 

Use an architect’s soft eraser for dry cleaning, or, if the dirt is too 
deeply ingrained, gently rub the surface of the print with a piece 
of stale (day-old) bread, using a hght circular motion and taking the 
greatest care not to raise the surface of the paper. Change the bread 
for a fresh piece when it shows signs of dirt. The back of the print 
should be cleaned as weU as the front. The print may be sterilized 
afterwards with thymol (p. 56). 

Dry cleaning may be all that is required. The use of organic sol- 
vents (petrol, &c.) is not recommended except for removing specific 
stains. If dr^^ cleaning in itself is not sufficient, two courses are open — 
either to clean the print as a whole by immersion in water, or to 
apply local treatment to remove specific stains. 

2. Immersion methods 

The print requires support when immersion methods are em- 
ployed. Some workers use a sheet of plate glass, others a flexible 
support of stiff paper or Polythene. The print is never hftedfrom the 
water by its comers or handled while wet. It is the support that is 
handled, and when the support is slowly raised out of the water the 
print rests upon it and adheres so that it may be safely moved and if 
necessary turned over on to another support without being strained 
in any way. 

A good soaking in cold water will always freshen up a print. After 
an hour it may be placed in a bath of hot water. Most fly-marks and 
mildew stains will respond to such treatment alone. 

3. Cleaning with soap 

In cases where further general cleaning is necessary, soap may 
prove to be a satisfactory cleaning agent. Carefully test the effect in 
a comer first. Lay the print face downwards on plate glass, damp it 
by contact with wet blotting-paper, then apply a httle good-quahty 
soap foam to the back with a large camel-hair mop. In larger prints 



a very soft badger shaving-brush is a convenient tool, and if the tips 
of the hairs only are used without pressure, there should be no strain 
on the surface of the damp paper. If the results are satisfactory on the 
back, the face of the print may be treated similarly. AU trace of soap 
must be removed afterwards by thorough washing; if this is omitted, 
the paper may go yellow. 

Alternatively the print may be soaked in a dilute solution of a non- 
ionic wetting-agent (e.g. Lissapoft or Nonex^) for half an hour and 
then washed. 

Washing is so important, especially when soap and chemicals have 
been used, that it must be carried out with something of the ritual 
of the photographic studio. 

Lead a slow current of water by a rubber tube to the bottom of 
the dish containing the print, and allow the water to emerge below the 
support. The duration of washing will depend on the nature of the 
foreign matter to he removed and on the type of print under treat- 
ment, but at least an hour should be allowed where bleaching solu- 
tions have been employed. 

4. Drying and removal of creases 

In order to dry the print, it is laid face downward on a pohshed 
glass plate, and pressed on the back with a pad of blotting-paper to 
remove excess moisture. It is then set up to dry slowly in a gentle 
draught of air. Contraction of the paper on drying against the glass 
win remove most creases and marks of folds (see also p. 85). 

If the print has lost its mechanical strength, it would be resized at 
this stage (p. 87). 

The processes that have been described so far will freshen up a 
print and in many cases this wall be all that is required. It is only when 
staining is persistent that it is necessary to have resort to chemical 
methods of cleaning — bleaching, and the action of specific solvents. 

’ I.C.I. Ltd., supplied by Hopkins and Williams, Freshwater Rd., Chadwell 
Heath, Essex. 

^ Gemec Chemicals Co., 120 Moorgate, E.C.2. 




Stains that have survived ordinary washing may often be removed 
by the process known as bleaching. This involves either treating the 
stain with a substance that wiU break down the colouring matter by 
oxidation into simpler colourless compounds that may be washed 
away, or treating the stain with a substance that will reduce it to a 
colourless compoimd that remains in situ. The most effective and 
permanent bleaching agents are in the first category (oxidizing 
bleaches) : chlorine dioxide, hypochlorites, sodium perborate, hydro- 
gen peroxide, and potassium permanganate. Examples in the second 
category (reducing bleaches) are sodium hydrosulphite and sodium 
formaldehyde sulphoxydate. While reducing bleaches are often effec- 
tive in treating dye-stuff stains that withstand the action of oxidizing 
bleaches, they sometimes merely change the colour of the dye, but 
even if they go a stage farther and decolorize the stain completely, 
there is always the possibility that colour wiU eventuaUy reappear 
owing to the subsequent oxidizing action of the atmosphere. 

The simplest form of bleaching by oxidation is exposure to air and 
sunlight (or ultra-violet light) and this has given remarkably success- 
ful results in the case of Egyptian papyri where the writing material 
is a stable carbon ink. 

The danger in using bleaching processes of any kind lies in the 
possibility of a loss of briUiancy in the inks and pigments and, if 
bleaching is overdone, the fabric of the paper may itself be attacked 
and weakened, especiaUy when hypochlorites are used. For this 
reason bleaching agents must be applied under strict control and for 
the minimum time necessary to achieve the required results, and any 
excess of reagent is then decomposed with chemicals or removed 
immediately afterwards by thorough washing with water. 

I. Use of hypochlorites 

The traditional method of bleaching paper depends on the action 
of chlorine generated by calcium or sodium hypochlorite. The cal- 
cium compound known as bleaching powder is largely employed 
for bleaching the raw materials from which paper is made, and while 



it may also be employed for bleaching prints, sodium hypochlorite 
is generally preferred as it is easier to prepare. Sodium hypochlorite 
is known, variously, as Eau de Javelle and ‘chlorinated soda’. The 
stock supply of sodium hypochlorite should be the commercial 
preparation marked ‘10% w/v available chlorine’ and it should be 
stored in a coloured bottle in a cool dark cupboard, otherwise it soon 
loses its strength. The solution must be diluted with water before use, 
generally one volume being added to twenty volumes of water; in 
no circumstances should the strength of the bleaching bath be greater 
than 6:20. The chlorine generated by this hquid has a powerful 
bleaching action, discharging the colour from dirt, mildew stains, 
fly-marks, and remains of varnish. If the action is prolonged un- 
necessarily, the paper itself will become a staring white and there is 
evidence that the cellulose may be degenerated and permanently 
weakened. The solution, which has an alkaline reaction, tends to soften 
the paper, but softening may be mitigated to some extent by having 
at hand a bath acidified with hydrochloric acid into which the print 
is transferred for a few moments now and then during the bleaching. 
Such a bath may be prepared by adding about a teaspoonful of con- 
centrated hydrochloric acid to a quart of water. As in aU processes 
involving the manipulation of wet paper, the use of a back-sheet or 
support is essential. 

During the bleaching process, any writing in iron-gaU inks will 
disappear unless protected beforehand. This is done while the paper 
is stfll dry, by a local apphcation of a solution of nitrocellulose 
(5 per cent.) in a mixed solvent consisting of equal volumes of acetone 
and amyl acetate. The nitrocellulose can be removed by a wash of 
acetone at the conclusion of operations when the print is dry. 

Bleaching is only allowed to proceed until stains become faint. 
The print is then passed through a bath of sodium thiosulphate 
(photographic hypo) of 2 per cent, strength to remove the residual 
chlorine, and washed thoroughly, when it wfll be found that the 
remaining marks gradually disappear. The hypo functions as an 
‘antichlor’ and its employment is always desirable when hypo- 
chlorites have been used for bleaching. 



2. Use of chloramine-T 

Chloramine-T is a much milder form of bleaching agent than the 
hypochlorites mentioned above and its use is recommended next 
to that of sodium chlorite for routine bleaching. Chloramine-T 
possesses the unique advantage that, when apphed to a print, its 
bleaching properties are soon lost and nothing of a corrosive nature 
remains on the paper. Washing may thus be reduced to a mini- 
mum or entirely dispensed wdth. The process is particularly suitable 
for water-colour drawings, coloured subjects generally, and drawings 
in bistre and sepia, as the reagent may be appHed locally to those 
parts of the work that are stained without exposing the whole work 
to the action of the bleaching reagent. 

Chloramine-T is obtainable commercially in the form of a fine white 
moderately soluble powder and must be kept in a well-stoppered 
bottle on account of its instabihty. Dissolve, only immediately before 
use, 2 gm. in every lOO ml. of water. Apply the solution to the stain 
with a soft camel-hair brush, cover with a pad of blotting-paper, and 
place under a sheet of glass. After an hour the print should be ex- 
amined. Further applications may be necessary since the reagent is 
very mild in its action. 

3. Use of sodium chlorite [chlorine dioxide) 

This is the safest bleaching agent to use as there is no chance of its 
chlorinating the cellulose fibres of the paper; it is effective and easy 
to use but requires speciahzed equipment. The simplest method, 
which may be apphed where there are fume cupboard facihties, has 
been described on p. 60. 

4. Bleaching coloured and delicate prints 

Many prints and drawings cannot be immersed in hypochlorite 
solution without damage, notably when colours are present. In the 
case of a stained coloured print, the best that can be done is to lay 
the print face down on glass; place a pad of wet blotting-paper on 
the back of the stain for a few minutes and then follow this with the 



minimum of bleaching solution which is apphed to the moistened 
back of the print with a camel-hair brush. It permeates the paper, 
having a rmld bleaching action on the front. The stain should be kept 
under observation through the glass. When the result is satisfactory, 
flood with 2 per cent, hypo solution, then wash the print in situ, 
holding it against the glass all the time and at an angle to the running 
water. Now reverse the print, using a flexible support, and lay the 
support against the glass with the print face upwards, allowing a 
gentle stream of water to run over the picture for a short time. The 
print is then allowed to dry in this position. 

An alternative method of apphcation has been suggested for clean- 
ing India-proofs and very thin papers. A sheet of blotting-paper is 
impregnated with a rather stronger solution of chlorinated soda, say 
4 : 10, and allowed to become almost dry before being folded around 
the print and then sandwiched between two sheets of plate glass. 
The greatest care must be taken when removing the blotting-paper, 
and if there is any tendency for it to stick, it must be damped, other- 
wise it may bring away some of the ink and damage the print. When 
prints are of such a nature that they cannot be washed after this treat- 
ment, they should be exposed freely to the air for at least 2 days 
before being returned to the collection. It should be noted that where 
‘antichlor’ cannot be employed, the method is not without danger 
to the paper, even though, judging by appearance, the result of 
operations seems to be entirely satisfactory. 

5. Use of specific bleaching-agents and solvents 

(a) Oil, fat, and tar stains: Pyridine. Pyridine, only the purest form of 
which should be employed, is an invaluable solvent for old partially 
oxidized oil, and for asphaltic stains, being decidedly more effective 
than benzene. 

(b) Wax and candle-grease stains: Petrol. Some of the grease can 
generally be removed with a paper-knife. The whole print is then 
immersed in a bath of petrol. After soaking a few minutes the stain 
is rubbed gently with a camel-hair brush and soon disappears. 

(c) Fly stains: Hydrogen peroxide, &c. Stippling the spots with 


hydrogen, peroxide in an equal volume of alcohol is often effective. 
If this fads, try stippling with 2 per cent, aqueous chloramine-T. 

(d) Tea and coffee stains: Potassium perborate. Damp the area. Stipple 
a 2 per cent, aqueous solution of potassium perborate on the stain 
and expose to surdight for an hour or so. The bleaching action is slow 
and as the reagent is adcaline it is not without danger to the paper. If 
the paper seems to be softened unduly, the action should be stopped 
at once by flooding the affected part with water. The final bleaching 
may be done by using ethereal hydrogen peroxide after the paper 
has been ado wed to dry' (see (J) below). 

(c) Ink stains. Owing to the great differences in iron-gad inks, and 
even in modem inks of the blue-black type, no single process can be 
advocated as certain of success. A number of methods are avadable. 
Some of these may be found to bleach most of the stain, leaving a 
yedow tinge on the paper which, in turn, may be discharged by an 
entirely different process and reagent. Some of the possibdities are 
detaded below: 

Brush with freshly prepared 2 per cent, aqueous chloramine-T. 
If not completely effective after two or three appHcations, and if the 
subject adows, try 5 per cent. oxaHc acid or 10 per cent, citric acid, 
and then wash thoroughly. No tannate of iron can survive any one 
of these methods without being bleached, and if coloured matter 
remains, it is likely to be carbon or the residue of some dye-stuff. 
Always stop the action short of complete bleaching, and take no risks 
as regards the possibdity of rubbing up the surface of the paper. 

Should a coloured stain stid persist when dry, a stronger bleaching 
agent wid be required. Damp the stain and cover with powdered 
sodium formaldehyde sulphoxylate. This bleaches most ink stains, 
iron stains, and many dye-stuffs used in coloured inks, but thorough 
washing is essential after such treatment. 

One further method may be mentioned, strictly as a reserve pro- 
cess to be used in case of emergency. It must be regarded as a last 
resort, as it is definitely deleterious to the paper; at the same time it 
seldom fads. An aqueous solution (0-5 per cent.) of potassium per- 
manganate is painted over the stain where it forms a brownish-red 



blotch of manganese dioxide. After about five minutes cover this 
with a 2 per cent, aqueous solution of oxahc acid. The brownish-red 
colour soon disappears, and the paper will usually be found to bear 
no trace of the original stain. Thorough washing after treatment is 

(J) Blackened luhite lead and red lead: Hydrogen peroxide. White lead 
(basic lead carbonate) is readily converted to the black sulphide by 
the action of sulphuretted hydrogen gas present in industrial atmo- 
spheres. Red lead, an oxide, is similarly affected and in each case 
stains are caused which are very disfiguring. The black sulphide may 
be readily oxidized by hydrogen peroxide to lead sulphate which is 
white and thus it is possible by simple treatment to remove the 
staining and restore the brilhancy of the white pigment. In the case 
of red lead that is superficially blackened, treatment by hydrogen 
peroxide results in the formation of a thin white veil of lead sulphate 
covering, but not concealing, the red pigment, and as it is unusual for 
red lead to be converted more than superficially to sulphide, the 
hydrogen peroxide treatment is generally all that is required to effect 

The direct appUcation of a commercial solution of hydrogen 
peroxide is sometimes recommended, but this is undesirable as it is 
hable to contain corrosive impurities. The following methods are 
designed in order to prevent contact of the corrosive impurities with 
the paper. The hydrogen peroxide solution should be poured on a 
stucco plate or porous tde, which is then fixed over the print about 
^ inch above it and left for a few hours. The vapour wtU clear the 
blackened pigments. Similar results are obtained by using a shallow 
bleaching-box with a false bottom of stucco that can be moistened 
with hydrogen peroxide. The print is placed in the box, face upwards, 
and left there until the stain has disappeared. 

A method of more general apphcation is to employ an ethereal 
solution of hydrogen peroxide prepared as follows. Equal parts of 
hydrogen peroxide (20 vols.) and ether^ are shaken together in a 

* Ether must be used only in a well-ventilated room free from naked lights as it is 
highly inflammable. 


' »;- St 

(4) Dark res, due of rcsiu obtaured by cvaporacng the aleohoUc extract 
(s) The book, rebound 

t 1 . 1> FU N T I D 

book (ITALIAN 

i 6TH CtNT.) 


glass-stoppered bottle. The liquids are immiscible — the aqueous layer 
containing any impurities remains at the bottom, the ether layer rises 
to the top and this contains sufficient hydrogen peroxide for bleach- 
ing purposes. A glass tube (diameter ca. J inch) plugged with cotton- 
wool may be used for dabbing on the reagent. The tube is held in the 
hand, the cotton-wool dipped into the top layer only, and the ethereal 
solution is then apphed to the stained areas by dabbing. Lead sulphide 
stains are bleached very readily by this method (Pi. 12 A ands). The 
action of the vapour may be prolonged by having a piece of blotting- 
paper at hand to shp over the print immediately after applying the 
reagent. The bleaching action may be intensified, if necessary, by 
adding a drop of perhydrol to the ether. 

It is not unusual to find painted miniatures on paper or parchment 
heavily stained in places where white or red lead has been used, the 
stains being black or deep brown and having sometimes a silvery 
sheen. These can be freshened up very easily with hydrogen peroxide, 
and for this type of work it is best to employ the ethereal solution 
which may be applied locally with a small brush. If there is no 
immediate response to the action of hydrogen peroxide, it may be 
that the stains have arisen from the action of hydrogen sulphide on 
metallic silver. Silver was often used in Persian mimature landscapes 
to represent water, and when once blackened by sulphide, no satis- 
factory method can be recommended for its restoration. 

Occasionally, a black smudge of lead sulphide will be found un- 
expectedly on a drawing, and closer inspection will reveal it to be a 
residue firom a previous restoration that has been carried out either 
in white lead or in coloured pigments that have been mixed with 
white lead. In such cases the hydrogen peroxide treatment works 
satisfactorily as, even in the case of a pigment mixture, the original 
hue is restored by the conversion of lead sulphide to the white 

After hydrogen peroxide treatment there is much less chance of 
the white lead being blackened in future, because the sulphate is less 
susceptible to change than the basic carbonate. 

B 8167 





1. Japanese prints 

Few of the ordinary methods of treatment apply in the case of 
Japanese prints on account of the soft texture and quahty of the 
paper. It is fortunate that this paper does not appear to be greatly 
attacked by foxing or mildew and the dirt is generally superficial. 
To clean a Japanese print lay it face down on glass and cover it with 
a rather larger sheet of tissue-paper (or preferably Japanese tissue 
which is thinner), in such a manner that one end of the tissue projects 
and can be held against the glass. Apply plain water aU over the 
tissue in a series of Hght parallel strokes using a tuft of cotton-wool; 
sufficient will soak through the tissue to make the dust on the back 
of the print adhere and on carefully folding back the tissue this dirt 
win be removed. If not entirely successful, a very dilute gelatine solu- 
tion (half strength, see page 87) may be substituted for the water, a 
fresh operation being conducted on the same hnes. After cleaning the 
back treat the front similarly. As mauve and heliotrope pigments on 
Japanese prints are generally of vegetable origin and easily damaged, 
they should never be damped. 

2. Japanese vellum mounts 

The so-called Japanese vellum forms an ideal mount for Japanese 
prints. Although it is a hard paper, the surface will not bear cleaning 
by the ordinary dry methods as it is very easily rubbed up. A clean 
soft sponge or, preferably, a tuft of cotton-wool is dipped in dilute 
starch solution, squeezed to remove excess Hquid, and passed Hghtly 
and rapidly over the surface. By this means the dirt is picked up and 
the mount greatly improved in appearance. 

3. India proofs 

The cleaning of an India proof is an extremely dehcate matter on 
account of the ease with which the India paper becomes detached 
from its mount. When once detached the difference in expansion and 
contraction between the thin paper and its thicker mount renders it 
almost impossible to regain the original tension between the two 


when they are remounted and dried. It should be noted also that the 
sharpness of detail in an India proof is bound to be lost to some 
extent if it is necessary to wet the paper for cleaning. Stains should 
therefore be removed, for preference, by dry cleaning or by local 
treatment if possible. 

Reference has akeady been made (p. 78) to the method of clean- 
ing using blotting-paper impregnated with chlorinated soda. This 
method, as has been suggested, is open to criticism, but alternative 
methods milder in action would necessitate longer contact with the 
damp bleaching-paper and result almost inevitably in the separation 
of the print from its mount. If the two sheets are only partially 
separated, they had best be completely detached from each other by 
soaking and floating them apart, and then chlorine dioxide (p. 60) 
or a chloramine-T bath may be used to clean both the print and the 

In order to remount the dry print it is laid face down on plate 
glass; the back is pasted all over, taking care to keep the paste off 
the glass, and the glass is then turned over and placed against the 
mounting paper which has been previously damped so that the India 
proof adheres in its correct position. The mounted India proof is 
dried superficially with blotting-paper, and after about ten minutes 
removed ftom the glass. It is then allowed to dry slowly under a 
weight protected by sheets of blotting-paper on each side. Should the 
blotting-paper adhere, it may easily be removed by damping with 

For the preparation and use of flour paste see Appendix III. 

4. Pastel and chalk drawings 

Pastel dratvings, especially those on vellum, should be ftequently 
inspected as they are liable to become infected with moulds. Moulds 
may attack certain pigments (Indian yellow; bile yellows) or they 
may grow on the binding medium of the powder colour — generally 
gum tragacanth — or on the sizing of the paper. 

When mildew is foimd growing on a pastel it must be picked off 
with a fine camel-hak brush shghtly moistened with pure alcohol. 



The picture is then steriHzed with thymol. If the mould growth has 
caused permanent staining it may be possible to hide this by spread- 
ing the pastel pigment with a stump. Pastels should always be kept 
in a dry place and they should be framed in contact with a sheet of 
paper impregnated with thymol or Santobrite (see p. 57). 

The treatment of pastels by amateurs should be restricted to the 

When certain types of chalk drawings (outline drawings on in- 
ferior brown paper) are mildewed they may become stained in a 
manner which is very disfiguring owing to the sparsity of pigment. 
The stains cannot be bleached as the paper would be bleached also. 
There is a tendency in old pastel and chalk drawings for the colour 
to be fixed in the paper as a result of the prolonged effect of traces of 
moisture upon the gum of the pastel and size of the paper; these 
adhesives bind the particles of pigment that are in direct contact with 
the cellulose tissue. Water may actually be used for cleaning drawings 
of this nature, as follows: float the drawing on the surface of cold 
water face upwards (with a paper support immersed below), and rub 
the spots gently after some minutes with a fine brush. If this is only 
partially effective after ten minutes, remove the drawing, using the 
support, and float it on to very hot water in the same way. This 
treatment is only possible with old drawings when the pastel pigment 
has become fixed in the tissue of the paper. The mould growths are 
mainly on the surface of the paper and are removable after softening 
on water. By this treatment the worst of the staining is dispersed. 

5. Bound papers 

It is sometimes necessary to remove a page from a book for treat- 
ment by one or other of the methods described above. This is done as 
follows — insert a wet string against the page near its hinge and close 
the book. After a minute the page may be tom out. When the page 
is ready to be returned, the tom edge is pasted and the page ahgned 
with the fore-edge and head end of adjacent pages; this wfll bring 
the pasted edge automatically into register with the hinge from which 
it was tom and it is caused to adhere by placing the book in the press. 




I. Removing creases from paper. When not vety bad the creases should 
he damped and a warm iron should then suffice to remove them. 

Reference has been made above to the removal of creases bv 


drying the print against glass. In the case of very bad creases the print 
is damped and laid face downwards on glass. Strips of soft white 
paper are then pasted aU round f inch on the glass and j inch on the 
print, so that the tension may be greater as the print shrinks on dry- 
ing. When the print is dry and flat, a little paste should be rubbed 
along the line of the crease and, if necessary, a strip of smular or 
thinner paper appHed as a patch and allowed to dry before releasing 
the print from the glass. In carrying out such operations the guard 
papers must not be heavier or stronger than the paper to be flattened, 
otherwise tearing may occur. 

2. Removing creases from parchment or vellum. With parchment or 
vellum the problem of creasing and warping is much more com- 
plicated and even prolonged treatment in a press is seldom effective. 

A practical method of flattening parchment is first to relax it by 
sandwiching the membrane between two sheets of damp blotting- 
paper and pressing Ughtly for a time. The blotting-paper should be 
uniformly damp but not too wet. The time required for this opera- 
tion will be determined by the nature and thickness of the parchment. 
When the membrane is limp, it is placed on a glass that has been 
previously poHshed with French chalk to remove grease, covered 
with a sheet of dry blotting-paper of the same size, and held against 
the glass by weights placed around the edges. Strips of ^inch lead 
are ideal for the purpose and they may be as broad as a rifler. Weights 
should not be so heavy as to immobihze the parchment, the process 
requiring that, as the parchment dries and contracts, it will sHde over 
the glass and take the weights with it, the centripetal puU being 
rather greater than the restraining action of the weights. By this 
means, even long-standing creases may be made to disappear. The 
most difficult types of membrane are those cut from near the edge 
of the skin, where thickness and grain are incHned to be variable, and 



these may need localized treatment in some form of stretching frame 
in which the forces can be adjusted to the requirements of the mem- 
brane. Such an apparatus has been described by Cockerell^ and there 
are cases in which the frame is an invaluable adjunct in removing 
creases from parchment. 

If such operations are practised on worthless documents it will be 
noticed that care is necessary to prevent ink spreading and it is a wise 
precaution to fix ink beforehand with nitrocellulose as mentioned on 
p. 65. When colour is present, and especially gold lettering, the 
membrane must on no account be relaxed. Restorations of this nature 
can only be conducted satisfactorily by an experienced bookbinder. 

3. Tears in paper. If the quaUty of the work allows, immerse the 
paper in water face downwards on a glass support, and when the 
tom pieces have been carefully floated into their correct positions 
slowly raise the support and paper out of the water. When half dry, 
tap along the joints with the back of a spoon in order to weld the 
fractured surfaces together. Paste may be added as a reinforcement 
and also a patch if necessary. 

When machine-made paper is being repaired, the grain of the 
patch must be ahgned with that of the paper. The direction of the 
grain of a machine-made paper may be discovered by moistening 
two edges at right angles; the paper swells in the direction across the 
grain, and acquires a crinkled appearance along the edge. 

Hand-made paper should be patched with another piece of thiimer 
hand-made paper using rather dry paste. Press very hghdy there- 

When an insertion is required, it should be of the same age, thick- 
ness, and surface appearance as the paper being repaired. The restorer 
must have at hand a reserve collection of sample papers, engravings, 
&c., from which to select repair material when necessary. Even when 
the greatest care is taken in matching the papers, some allowance may 
have to be made for variations of sizing, and differences of shrinkage 
on drying. The tendency is to make the patch too damp with paste. 
T he edges of the insertion and the paper are chamfered as described 
* Cockerell, D., Bookbinding and the Care of Books, pp. 329 seq. 1901. 


below in the repair of parchment, and after the edges have been 
pasted and the insertion stuck in position the whole is flattened in the 

When a Japanese print is broken or much tom, it should be com- 
pletely pasted down on another Japanese paper of similar quaUty and 
strongly pressed. 

4. Tears in parchment and vellum. When creases are absent, there is no 
need to soak parchment or vellum. A simple chamfered tear may 
be repaired by sticking the edges together and pressuig under a weight 
tin dry. Paste, however, is not a very good adhesive for parchment. 
The replacement of tom comers requires a much more powerful 
adhesive, and stronger joints are obtained when the edges are painted 
with the mi nimum quantity of dilute acetic acid (say 10 per cent.). 
This ge1a tini7. es the fibres so that the tom edges can be pressed inti- 
mately together. 

Where there are no natural chamfers, as in a plain cut, the edges 
must be artificially chamfered by mbbing with an abrasive stick of 
chisel shape. This enlarges the cut, making it a V-shaped sht, and a 
patch of similar appearance but rather greater in size and thickness is 
cut to shape and then chamfered down to fit. It is inserted after 
applying adhesive to the edges, dried in the press, and then brought 
to the extact thickness by mbbing with a block of fine pumice. 
When such work is carried out carefully and the sheet rubbed over 
with a Httle pumice flour at the end of operations, the repair may be 
so perfect as to be ahnost beyond detection. 

5. Sizing and retouching. The paper will require to be sized afresh 
if any retouching has to be done. Parchment size is sometimes recom- 
mended for this purpose but it has no advantages over gelatine. A 
good gelatine size is made by dissolving one sheet of clear gelatine in 
a quart of water (not more than 1*5 grams per Htre). Size should 
always be fre shl y made immediately before use, and brushed thinly 
and Hghtly over the paper. Thick papers may be sized by immersion 
and hung on a line to dry away from radiators and draughts. 

If any retouching is necessary, this should not exceed the absolute 
minimum required to conceal disfigurements due to tears and 



abrasions of the paper. The method should generally be by stippling 
save where broken lines have to be joined up. 

6. Toning. When one page of a book has been bleached or when 
one of a uniform series of engravings has been similarly treated, it 
may appear blanched and uninteresting beside its fellows. This may 
be remedied in a rule-of-thumb fashion by staining the paper with 
stout or a decoction of tea or coffee. Preliminary trials are made on 
similar paper in order to decide which stain gives the best results and 
to determine a suitable concentration. 


I. Parchment and vellum 

Great care is required in mounting parchment and vellum as these 
are generally mounted in the damp relaxed condition and contrac- 
tion on drying is considerable. If the paste is not uniformly apphed 
around the edges, distortion will result. Assuming that there are no 
illuminations present, the parchment is relaxed in water and placed 
in sheets of dry blotting-paper between glass plates until it appears 
almost dry, while still retaining its suppleness. A thick moimting 
board is selected and four guide-marks made to show where the print 
is to be permanently fixed. The membrane is laid face down on glass 
and pasted half an inch all roimd the edge two or three times. It is 
then turned over and placed in its permanent position on the mount, 
and the edges pressed down very thoroughly with clean blotting- 
paper. It must be remembered that the edges will have to withstand 
a fair strain as the parchment shrinks on drying. Should the parch- 
ment have been too wet when mounted, the edges will pull away, 
dragging off part of the moimt with them. Before drying, remove 
any marks with a damp sponge. After covering with two or three 
sheets of blotting-paper, weights are apphed over the four comers 
and the membrane is left to dry, when it will become as tight and 
smooth as a drum. 

Vellum is treated in similar fashion. The thinnest and finest speci- 
mens do not require prehminary soaking. It is sufficient to leave them 


for a short time in damp blotting-paper pressed between glass plates 
before mounting. 

When there is colour or gilding present, the membrane must on no 
account be relaxed, and manipulative processes are best left for the 

2. Papers 

In the absence of special laminating equipment (p. 63), manu- 
script material is commonly mounted in guard-books (Pi. 8). 

Prints and drawings are never laid down’ (pasted down aU over) 
on card, but attached to the card by hinges of thin paper folded back 
on themselves and known as guards. When only one such hinge is 
used the drawing can be turned over to inspect the reverse and water- 
marks can be studied by transmitted hght. More usually four guard 
papers are employed and there is then less chance of accident by care- 
less handling. 

Only the best white or ivory cards should be used for mounts. 
These are cut to size and increased in thickness as required by pasting 
cards together. The mounts should be strongly pressed after pasting. 

AU mounts should be finished with rounded comers to avoid the 
chance of damage should they be accidentaUy dragged over the sur- 
face of adjacent prints. This can be effected by smoothing the edges 
and comers with glass-paper. 

Three types of mount wiU be described — the sohd, the overthrow, 
and the window mount. These are Ulustrated in Fig. 4 (p. 90) . 

{a) The solid mount. This type of mount is the one most commonly 
used, when there is nothing of interest on the back of the print. The 
print or drawing is attached by pasted paper hinges in a ‘box’ pre- 
pared by pasting a cardboard (3 sheet) frame to a backboard (4 
sheet) of the same size as the frame. In the sohd type of mount the 
cardboard frame does not cover any of the print. The backboard 
must always be thicker than the frame, the relative thickness depend- 
ing on the size of the print. A comparatively large print would re- 
quire a strong backboard for support. Deep ‘boxes’ must always be 
employed in mounting pastels. 



(b) The overthrow mount. This type of mount is in general use where 
prints are hinged by one guard paper only so that the back can be 
inspected. It is also used when a print has a tom edge. Such a print 
is mounted on a backboard by pasted guards of paper and a frame is 


prepared of such a size as to conceal the damage. This frame is 
hinged to the backboard by a strip of Hnen so that it may be opened 
to the left to expose the whole of the print for examination. A hinged 
arrangement of this kind is known as an overthrow mount. 

(c) The window mount. When drawings occur on both sides of the 
sheet of paper the paper may sometimes be spht so that the drawings 
are released and they may then be mounted separately. As the 
operation of sphtting paper is always attended by risk, it is generally 
preferable to adopt instead the window type of mount so that both 
picttures remain visible. In this case the paper is attached only by its 
edges between superimposed rectangular openings of appropriate 
size in two cards. As the drawing might easily be damaged if left 
unsupported it is usual to mount a sheet of cellulose acetate^ at one 
side so that it forms a transparent support. If the paper is very frail 
or where much ink is present that may have rotted the tissue, it may 
be necessary to give it additional support by adding a sunilar sheet 
at the other side. 

It is a general principle in mounting a collection to endeavour to 
make mounts in a limited number of sizes as this is convenient both 
for storage and for exhibition purposes. The card surround should be 
of such a thickness as to protect the picture from contact with the 
glass when framed. If the picture is in contact with the glass, con- 
densation effects may encourage the growth of mildew. 

3 . Documents with Seals 

When documents bearing appended seals are to be framed, they 
should first be inlaid in a sohd card mount of such a size and thickness 
that the seals are backed by, and nested in, the card, so that their 
weight is supported. This protects them from the possibihty of acci- 
dent should the fabric by which the seals are attached become 
tendered and no longer able to sustain their weight. 

W^ ax seals are easily broken and, if not repaired at the time, frag- 
ments tend to get lost and reconstructions become more comphcated. 
A mixture of equal parts of rosin and beeswax is usually recommended 
‘ e.g. Celastoid. British Celanese, Ltd., Hanover Square, London, W.i. 



as an adhesive for wax and, if it is a case of dealing with a simple 
fracture, this adhesive may be spread in the form of a hot hquid on 
the joints, pressing them together immediately rill the parts are held 
securely. With heavy seals it may be necessary to insert one or two 
dowels before m akin g the joint. The holes for these are made by 
probing with a hot needle, and the actual dowels can be made from 
fragments of broken needles of appropriate length and thickness. It 
is sometimes possible to reinforce a heavy seal by attaching a thin 
adhesive tape around the circumference in such a manner as to be 
tmobtrusive, and it may be painted thereafter to match the colour of 
the wax. 

When a seal is sphntered and parts are missing, the main fragments 
should be brought together as above and any lacunae made up with 
plaster of Paris containing a Httle ochre and black as a rough match 
for the natural wax colour. Green and red colours are sometimes re- 
quired. The tinted plaster should extend sHghtly above surface level, 
and after it is quite dry it is carved to match the design. It is then im- 
pregnated with thin celluloid lacquer to stop the suction and painted 
to match the original. While a repair executed in this manner should 
not be obtrusive or out of colour harmony, there should be no effort 
to conceal the fact that the seal has been restored. 



The spinning of threads and the weaving of textiles come very early 
in the history of craftsmanship. Linen in southern Europe can be 
dated back as far as the Stone Age; in northern Europe wool was in 
use in the Bronze Age, and sdk, which originated in China, can be 
traced back for more than 5,000 years. Embroidery, i.e. the decora- 
tion of the finished textiles, did not appear till much later. 

Natural fibres may be of animal origin (wool, silk), or of vegetable 
origin (cotton, Hnen, hemp). Animal fibres contain keratin, a nitro- 
genous compound, and, when burnt, they contract and smell of burnt 
feathers due to the decomposition of the keratin. Vegetable fibres 
are composed of cellulose; they bum easily and give off a character- 
istic odour of burnt paper. Thus, it is possible by a burning test to 
make a broad distinction between animal and vegetable fibres, but, 
for specific identification, individual fibres must be examined under 
a microscope at about 100/150 diameters magnification when, with 
a httle practice, the main types can readily be distinguished. ^ 

As textiles are of an organic nature and subject to attack by moulds 
and bacteria, the commonest factors that promote decay are those 
that favour the growth of these organisms, viz. a damp heat, lack of 
ventilation, and contact with decaying animal and vegetable matter. 
However, even under such unfavourable conditions ancient textiles 
may not be entirely destroyed. It is common, for example, to find 
fragments of textile surviving when they have been in contact with 
corroding copper, the corrosion products having acted as sterihzing 
agents. Textile remains are often found attached to the patina of 

‘ Cf. Identification of Textile Materials, 3rd edition, 1951, published by The Textile 
Institute, 10 Blackfriars St., Manchester 3. 


Chinese bronzes, in such a condition that the fibres can be identified 
and the weave recorded. Damp causes vegetable fibres to sw^ and 
soften, but the fact that animal fibres are not necessarily destroyed by 
immersion in water is proved by the survival of wooUen material in 
peat bogs and lake dwellings. * Excessive heat causes desiccation and 
embrittlement, and exposure to bright hght and noxious gases cause 
the type of deterioration known as tendering. The astonishingly 
fresh appearance of some Egyptian mununy wrappings may be due 
as much to the presence of salt as to the absence of hght and dryness 
of the surroundings. 


Before beginning restoration work on textiles the first step is to 
carry out an examination with a pocket lens or binocular naicro- 
scope, making a record of the following particulars. 

The nature of the fibres of both warp and weft threads. 

In spun threads, the direction of twist — S (') or Z ('). 

The type of weaving — tabby, twill, figured. 

The count per centimetre of both warp and weft threads. 

The presence of selvages. 

Dye-stuffs or apphed decoration — paint, gold leaf, metal threads, or 

The presence of sewing or stitch holes. 

(If the material is very dirty, making analysis difficult, gentle blasts 
of air from a blow-baU or bellows may be used to remove loose dust, 
or in some cases a soft brush may be used.) It may be impossible 
to record all particulars until the textiles have been washed, but it 
is important to get as much information as possible in the first place 
in case any evidence should be lost during treatment. The prehminary 
examination will decide what are the safest methods to adopt for 
cleaning and restoration. 

' Henshall, A. S., Miss, and Maxwell, S., ‘Clothing and other Articles from a late 17th- 
Cent. Grave at Gunnister, Shetland’. R. Soc. Proc. (Scot.) 1951-2, 86, pp. 30-42. This 
grave was found in peat, and when excavated, scarcely anything of the body was left, but 
all the woollen clothing in which it had been buried was well preserved. This clothing is 
now exhibited in the National Museum of Antiquities of Scotland, Edinburgh. 




Textiles are easily contaminated with soot and dust, and being 
absorbent they are readily stained by contact with substances in solu- 
tion, or decaying organic matter, or by the coloured substances pro- 
duced by moulds in the course of their growth. Material that is dirty 
or stained, but which is sufficiently strong to be manipulated, may be 
washed or dry-cleaned, and specific stains can be dealt with locally. 
It must be understood, however, that the processes recommended 
may require modification depending on the condition and type of 
textile that has to be treated, whether carpets, tapestry, lace, &c. 
Comphcations may also arise from the presence of a dye, or of gold 
thread, or because of the vagaries of the weave or make-up of the 

I. Washing 

The first requirement for washing textiles is a supply of soft water. 
Rain water or distilled water may be used, or water that has been 
sofrened (demineralized) by natural zeoHtes or by special synthetic 
ion-exchange resins. Old textiles are washed preferably in flat vessels. 
Small samples may be conveniently dealt with in photographic 
developing dishes, but for washing large textiles the most practical 
arrangement is to use a wooden tray lined with Polythene. Such a 
tray is constructed with sides about 4 inches high, in one of which a 
V-cut is made for emptying; a sheet of Polythene is spread in the 
tray and made to take the shape of the vessel, and is then held in posi- 
tion around the edges by spring clothes-pegs. By this means a water- 
proof tank of any size can be readily improvised. A very large tank 
may be emptied by siphon, or preferably hinged to the table along 
one of its longer edges, so that the water can be run out into an 
adjacent sink by tipping the vessel, the action being controlled by a 
long screw such as that commonly used for opening and closing 
hinged window-Hghts. In emptying the tank, the Polythene is pressed 
into the V-cut so that it forms a gutter between the table and sink. 

A loose sheet of Polythene should be placed in the bottom of the 



tank to act as a support for the textiles when they are hfted out of 
the water. 

In the case of textiles of open weave, drawn-thread work, or lace, 
these should be pinned down or tacked to a sheet of Polythene, for 
safety during washing. 

If textiles are coloured, a spot test with water will have to be 
apphed to each colour before washing to determine whether the 
colours are fast. Fugitive colours may be fixed either by treatment 
with a 5 per cent, solution of common salt or acetic acid of the same 
strength. A stronger solution of acetic acid up to about 20 per cent, 
may be used if necessary. A spot test should be made, and the spot 
absorbed in white blotting-paper to ascertain whether the salt or the 
acid fixes the colour. The whole textile should then be immersed in 
the appropriate fixing solution. 

When everything is in order, the tank is filled with soft water to a 
depth of about 3 inches, and the textile lowered gently into it, keep- 
ing the material flat and well spread out; it is kept immersed for an 
hour, changing the water as required, perhaps every twenty minutes. 
If the mains supply of water is soft, a gentle stream of running water 
may be used. This is conducted to the tank by a rubber or Polythene 
tube and .allowed to emerge beneath the Polythene support. When 
the running-water method of washing is adopted (in the case of a 
tank larger than the sink), a siphon draining system will be required. 

Some of the dirt will be soluble in water, and insoluble sandy 
matter may be released by gently tamping with the fingers after the 
textile has been soaking for some time. When the washing is finished, 
the water is run off, and the Polythene support, to which the textile 
adheres, is raised and maintained at an angle to allow the material to 
drain. W^hile still on the support the textile is pressed gently with 
absorbent towelhng and it is then transferred to a backing of warm 
absorbent material — towelling or flannel. 

W^hen half-dry, the textile is turned over on to a Polythene sheet 
and straightened out, the weave being adjusted so that the warp and 
weft threads are at right angles to each other. Fine brass pins are 
inserted vertically at intervals through the textile into the Polythene 

A. Photograph on Panchromatic Plate 

13. Photograph on Intra-Rcci Plate 


i_l. MANii’ii' or SI. ciniiiiiin irom Durham c ai' ii r d r a i 
AtU'r fliMninir ami inmiiitimr prim' to Ir.iniimr 



sheet. The pins may require adjusting from time to time as the 
textile dries. Drying should he carried out in a warm weU-ventilated 
room, and it may be facihtated by using a hair-dryer. 

Use ^detergents. While washing in plain water can bring about a 
remarkable improvement in the appearance of certain textiles, the 
use of a detergent is necessary if textiles are contaminated with 
grease. Soaps are not used as they tend to form an insoluble scum 
difficult to remove, but certain synthetic surface-active agents or 
wetting-agents may be employed. These are of two kinds, ionic and 
non-ionic, the latter type being the safer for textiles. Examples are 
* Lissapol-N^ and the preparation Igepal CA Extra.^ These substances 
can be used with confidence, but a warning is necessary against using 
any of the commercial cleansing powders and hquids that are based 
on them, because patent cleaners may contain in addition soap 
powder, soda, or other materials that might be deleterious to fine 
textiles. There is the added compfication that the formulation of any 
commercial product may be altered without changing the name and 
without any warning being given by the manufacturers. 

It cannot, of course, be assumed that wetting-agents are necessarily 
safe to use in the presence of dye-stuffs, and prehminary tests must 
always be made by applying a spot of the concentrated detergent to 
all the dyes on the material. If any of the dyes ‘bleed’ under this test, 
they must be fixed as described above with salt or dilute acetic acid. 

It is characteristic of wetting-agents that they are used in extremely 
dilute solution (as specified by the manufactinrers), quantities of the 
order of i fluid ounce to 5 or even 10 pints of water, I'.e. i_p?r cent. > 
to 0’5 per cent, strength, being adequate for most purposes. In using 
detergents, they should be added in the requisite amount to the water 
and thoroughly stirred so as to form a solution of umform strength 
before the textile is placed in contact with the hquid. During wash- 
ing, the wash-waters are changed at least three times, and a final bath 
of fresh soft water completes the operation. In cases where luke- 
warm water may be used, the cleaning will be even more effective, 

' Imperial Chemical Industries, Ltd., loc. cit. 

* Marketed by General Dyestuff Corporation, 435 Hudson Street, New York 14, U.S.A. 

B 6157 




but the water should never be used hot as this has a softening effect 
on some fibres. 

Use of sapqnins. Saponins are widely distributed in nature among 
the higher plants, and their virtues as substances having mild deter- 
gent properties have long been known. They are neutral in action, 
froth readily with water, and their cleansing properties are due to the 
ease with which they form emulsions with substances of a resinous 
or oily nature. Plant extracts containing saponins are largely em- 
ployed in the East for washing clothes. They do not seem to have 
any deleterious effect on dehcate fabrics, and are safer to use with 
coloured textiles than soap. The precaution should always be taken, 
however, to test the colours first. 

A cleansing fluid may be made by infusing Saponaria (Soapwort) 
with water; or saponin can be obtained commercially as a white 
powder. To use the powder a httle is switched to a froth with water 
and this is apphed with a soft brush of the shaving-brush type. To 
clean upholstery the furniture is supported upside down or at an 
angle so that when the froth has done its work and is soiled it will 
fall away from the textile rather than sink in. The froth is hghtly 
worked over the surface of the textile using only the tips of the 
bristles of the brush, excess froth being removed with a soft rag, and 
the textile dried thereafter with absorbent towelling. By this method 
excellent results are obtainable without rubbing and with the mini- 
mum of exposure to moisture. For cleaning damask that is frail with 
age, the process may be used with appropriate variations. It will 
usually be found that in cases where it is safe to use water, dirt can be 
removed more readily with the aid of saponin, and that colours are 
improved in appearance and enlivened by the treatment. 

2. Dry-cleaning 

Dry-cleaning may be carried out by mild suction, in conjunction 
with brushing, by using organic grease-solvents, or by dr)^ steam, 
the nature and size of the textile being the determining factors in 
deciding which procedure is the most suitable. 

Vacuum-treatment and brushing. Old carpets and curtains that are 



very dusty may be improved in appearance by brushing with a soft 
brush (in the direction of the pile) towards a vacuum cleaner, or, 
alternatively, a suction nozzle may be used if the fabric is strong 
enough to stand it. It may be added that it is sometimes possible to 
improve the appearance of a rug by gently rubbing fuller’s earth or 
even baking soda into the pile, leaving it there overnight and then 
extracting it next day with a suction nozzle. In dealing with old rugs 
care must be taken to preserve the direction of the pile; if this is upset 
by violent local cleaning, the distorted fibres will scatter reflected 
hght and the rug will have the appearance of being stained. 

Brushes should be selected with care to suit the job in hand. The 
bristles should be white for preference and generally long, and the 
degree of sofiness is also important. Stiff brushes do have their place 
among the cleaning equipment but should be used with reserve. 

Organic solvents. The term ‘dry-cleaning’ is usually employed to 
denote cleaning by anhydrous solvents. These are less hkely to affect 
any batter in the cloth than water, but as they may well have a greater 
action upon dye-stuffs it is essential that aU colours on a textile be 
tested individually by spotting with solvent before immersing the 
textile in the cleansing solution. Dry-cleaning is only given prefer- 
ence in cases where the presence of water is undesirable. The solvent 
most commonly employed is trichlorethylene (Westrosol), a non- 
inflammable hquid of high volatihty (B.P. 88° C.). The Ciba Experi- 
mental Dye-house Laboratories, Basle, recommend that it should be 
used pure and cold, and that textiles should be immersed for from ten 
to thfrty minutes at the outside. It is further recommended that where 
there is a danger of dyes running, dichlorethylene (B.P. 55° C.) should 
be selected as the cleaning agent in preference to trichlorethylene. 

It is well estabhshed that when garments are soiled with grease and 
perspiration they are more prone to be attacked by moth than when 
freshly laundered, and cleaning with solvent has the same effect as 
laundering in discouraging attack. It goes without saying that garments 
should always be cleaned before they are incorporated in the collection. 

Steam cleaning. For certain kinds of textiles, cleaning by steam (if 
properly controlled) may be much less drastic than either soaking wnth 



water or cleaning with organic solvents. Steam may sometimes be used 
for softening and removing stains even in the presence of fugitive 
colouring matter that would spread if wetted. A steam-gun apparatus 
that win provide wet or dry steam at option is now regarded as an 
essential in modem laundry practice, and it could be apphed with ad- 
vantage to cleaning certainkinds ofmuseum textiles — costumes, for ex- 
ample, or ethnographical specimens — as it makes possible a variety of 
cleaning operations that would not be warranted by any other process. 

3 . Removal of stains 

It is not always advisable to attempt the removal of stains from 
old textiles. Stains of long standing may have undergone a chemical 
change with the formation of insoluble matter that can only be 
discharged by bleaching, and this process would be likely to weaken 
still further an old textile that had already become tendered. 
In certain cases, however, the removal of stains is to be recom- 
mended. In the unfortunate event of a textile being stained by acci- 
dent, the staining should be dealt with quickly before it has had time 
to become fixed in the fibres. Iron stains commonly rot vegetable 
fibres, and these should be removed if the fabric is strong enough to 
withstand treatment. With frail old textiles, removal of stains may 
be a very dehcate operation requiring experience and some know- 
ledge of chemistry, but with modern fabrics, including many textiles 
in ethnographical collections, there is less risk of damage and, even 
for the amateur, a reasonable prospect of success. 

In the removal of stains the first step is to ascertain the nature of 
the textile (e.g. wool, silk, or linen), and it is also a help to know the 
nature of the substance responsible for causing the stain; this may 
be obvious, but sometimes it will be necessary to make tests before 
the appropriate reagent can be selected. * Even so, all sorts of difficulties 
are hable to be encountered. The textile itself may be decorated with 
paint susceptible to solvents; the colours of a dyed fabric may bleed 
in contact with liquids; or a particular fining may cause trouble by 
staining the material to which it is attached. In all cases the selected 
' Moss, A. J. E., Stain Removal, 1950 (Uiffe & Sons, Ltd.). 



reagent must be tested on an inconspicuous part of the textile to see 
whether it is safe to use before applying it to the stain. 

The success of the operation often depends on how the reagent 
is apphed. The general tendency is to use too much solvent and so 
risk spreading the stain. The smallest quantity of the reagent should 
be apphed on a small screwed-up tuft of cotton wool, and with a 
dextrous touch the stain should be hfted off rather than mbbed in. 
Several apphcations may be necessary, but if a stain is very stubborn, 
it is better to leave it rather than risk damaging the fabric. 

In certain cases a different technique is required, known as ‘ringing 
the stain’. It is obvious that the application of any organic solvent to, 
for example, a grease spot, would simply result in the grease being 
spread over a larger area of the textile. The procedure adopted here 
is to stretch the textile, stained side downwards, over a glass plate 
covered with blotting-paper and apply the reagent to the back in 
drops from a small pipette so that it forms a ring around the stain. In 
this way the solvent bears in upon the stain from aU sides at once, the 
grease is dissolved without spreading, and the resultant solution is 
absorbed in the blotting-paper. 

An alternative method for removing grease or wax stains is to 
cover both sides of the stain with clean blotting-paper and apply a 
warin^iron. This is the traditional method for removing ‘candle 
grease’; the grease melts and most of it runs into the warm paper. 
Any residue is then ehminated with benzene, turpentine, or trichlor- 
ethylene apphed by ringing as described above. A blob of hard- 
ened paint or a blob of grease or mud should be carefuUy removed as 
far as possible with a scapel or razor blade before the solvent is apphed. 

In using solvents for the removal of stains it must be borne in mind 
that animal fibres are very sensitive to hot water, which causes them 
to shrink and lose their lustre, to excessive rubbing, and to the use 
of those bleaching agents which depend for their effect on the evolu- 
tion of chlorine. Vegetable fibres are generally more robust, but 
after bleaching or using acidic reagents it is essential to wash the 
treated part of the textile until it is neutral. These observations are 
summarized below in Table I. 


Table I 


! Animal fibres 
(silk and wool) 

[ Vegetable fibres 
(cotton and linen) 

Strong acids or alkalis ' 

Never used 

Never used 

Dilute acids 

Never nitric acid 


Dilute alkalis 



Bleach containing chlorine 

Never used 

Use weak only 

Hydrogen peroxide 


Use weak only 

Very hot water 

Never used 

1 Permitted 

A list of specific reagents that are commonly employed for remov- 
ing stains is given in Table II which has been compiled with the object 
of systematizing procedure. The nature of the textile and of the stain 
being known, reagents are apphed to the stain one after the other in 
the order given; thus, for example, the symbol GLA for the removal 
of mud stains on silk is to be interpreted as follows — hydrogen 
peroxide (G) is appHed to the mud and, if no appreciable change is 
seen to take place in the course of a minute or so, very dilute am- 
monia (L) is applied. The stain should lessen in intensity and the mud 
be released so that it can be sponged off with hot water (A), taking 
care that it is not hot enough to destroy the lustre of the material. ^ 

Table II 

Procedure for the Removal of Stains 

List of Reagents 

A Soft water (Hot) 

B Methylated spirits 
C Spirit soap (B.P.C.) 

D Trichlorethylene 
E White spirit 

F Hydrochloric acid (2 per cent.) 

G Hydrogen peroxide (10 vols.) made 
alkaline with ammonia 
H Potassium permanganate (i per cent.) 

I Sodium hydrosulphite (5 per cent.) 
made alkaline with ammonia 

* Rubber gloves 

J Oxahc acid (i per cent.) 

K Acetic acid (0-5 per cent.) 

L Ammonia (o-i per cent.) 

M Sulphurous acid 
N Pyridine 
O Morpholine 

P Hydrofluoric acid* (i V0I.+3 vols. 

Q Chloramine-T (fresh, 2 per cent.) 

R Tartaric acid (fresh, 5 per cent.) 

should be used. 

‘ For more extensive treatment see Holden,]. T., and Fowler,]. N., The Technology of 
Washing (British Launderers’ Research Assoc.) and Moss, A.]. E. op. cit. 



Nature of stain 

Material stained 



Cotton or linen 

I. Mud .... 




2. Grease 


D or E 

D or E 

3. Iron rust 




4. Red ink 



Q or LB 

5. Blue-black ink 



Q or HMA 

6. Copying ink 




7. Marking ink 




8. Fresh oil-paint 

BCE or JA 



9. Old oil-paint 



N or OA 

10. Lipstick 





Textiles are liable to be presented for conservation in aU stages 
of deterioration, but perhaps the most difficult problem arises in the 
case of material that is fireshly excavated, possibly from a tomb. The 
textiles may be hard and brittle or so decayed as almost to resemble 
a spider’s web, or they may be damp and the pattern and colours 
concealed beneath a layer of dust and insect remains. But no matter 
what condition they are in, provided they have not completely dis- 
integrated, they will always repay careful examination, and it is sur- 
prising what may be done in the way of preservation. 

The first stage of treatment is always superficial cleaning, so as to 
enable the textile to be examined. Loose dust may be removed by 
bellows, but on no account must the material be brushed. Insect 
remains often adhere firmly and these have to be removed with 
scalpel or forceps working under a binocular microscope. If a textile 
has been damp, it may gradually develop a white bloom as it dries 
owing to the crystalhzation of salts, and this should be held in check 
by spraying^ occasionally with a fine mist of distilled water — ^not 
enough to soak the fabric and make it limp, but merely to prevent 

' Wien fatty acids are present the textile may be difficult to wet. If there is any diffi- 
culty in wetting the textile with the water spray, a Uttlc wetting-agent (Lissapol) should 
be added. 



the surface from becoming white and so obscuring the structure. 
When extraneous matter has been removed as far as possible, reverse 
the textile and deal with the back in the same manner. 

It is nearly always possible to wash frail textiles provided they are 
manipulated on a support and handled as Httle as possible while wet. 
As dirt and dust have a deleterious effect, washing is always desirable, 
but where there are decorative attachments, e.g. metal threads or 
leaf, leather, &c., that might be damaged by water, washing must be 
avoided. In such cases the textile should be photographed after pre- 
hminary cleaning, and by using an infra-red plate it may be possible 
to record details of ornament that are either concealed or that are 
difficult to interpret by ordinary photography (Pi. 13 A and b). 

Sometimes textiles are rigid and crumpled or folded. When in this 
condition they should be laid on a glass plate or Polythene sheet and 
the wrinkles and folds sprayed so that they gradually become re- 
laxed. If the material can be washed, no attempt should be made at 
this stage to stretch the fabric out. The material is floated into shape 
after it has been immersed in water. 

Details of the washing process have already been described. A 
modification in the case of frail textiles is to use a porous support^ 
instead of the Polythene sheet, so that water with dirt and sandy 
material can drain through. Alternatively, the material may be sand- 
wiched between two sheets of Terylene net fixed in a frame so that 
handling is reduced to a minimum. After washing, the Terylene is 
pressed between warm dry towelling to remove excess moisture, 
and when partly dry the frail material is transferred to a sheet of 
Polythene and pinned out as described on p. 96. 

The greatest advances in the techmque of handling ancient textiles 
have been made in the Textile Museum, Washington, D.C., and are 
fully described and illustrated in Workshop Notes pubhshed half- 
yearly by this institution. In order that the routine of cleaning may 
be visualized, a short summary of the operations that apply in the 

’ Mrs. F. S. Greene {vide infra) recommends Lumite, a stifF plastic material having an 
open cloth weave. An Enghsh equivalent is Lumarith manufactured hy British Celanesc 
Ltd., Hanover Square, London, W.i. 



case of woollen textiles is appended. For full details the reader is 
referred to Workshop Notes No. 1.^ 

1. Textile analysis and scrutiny. 

2. Test the chosen detergent on colours and fix them if necessary. 

3 . Support the textile between open-work screens. 

4. Degrease in Stoddard Solvent (a mineral spirit) and allow to dry 

5. Digest in a warm concentrated solution of Parazyme (a starch and 
protein digestive enzyme mixture), to render stains soluble, and 
flush with water at the same temperature. 

6. Bathe in water fortified with detergent. (Igepal-CA Extra is recom- 
mended, one part in two hundred parts of distflled water.) Tamp, 
if necessary, through the open-work screen to remove residual 
earthy matter, and finally rinse — the last rinse in 10 per cent, 

7. Dry at first through the screen, latterly, if practicable, with a paper 
handkerchief placed next the textile. 

8. Transfer to a plastic backing that will hold small blocking pins, 
and straighten the warps and wefts. 

9. Mount the dry cleaned textile on a thin cellulose acetate sheet 
by occasional stitches of thread of the colour of the warps, inserted 
from the back between two warps, and passing over only one warp 
so that the stitches are eventually concealed by the wefts. 

to. Frame the mounted textile passe-partout between two sheets of 
in. Plexiglass (Perspex). 

* Note: Woollens shrink and harden if exposed to changes of temperature while wet. 

In comparative experiments conducted at the British Museum 
Laboratory with Miss Louisa Bellinger, Curator-Analyst of the Tex- 
tile Museum, Washington, D.C., it was decided that Lissapol-N 
(l.C.I.) gave results comparable with those obtained in using the 
American preparation Igepal-CA Extra. Solvent Naphtha or white 
spirit may be used in place of Stoddard Solvent, and Diastase^ instead 
of Parazyme. 

’ Greene, Mrs. F. S., Cleaning and Mounting Procedures for Wool Textiles, Textile 
Museum, Washington, D.C. 

^ Cutrilin Ltd., 17 Victoria Street, London, S.W.i. The manufacturers recommend 




Even when in a fragmentary condition, textiles must never be 
stuck down with adhesives. These harden and damage fabrics, they 
are also nutrients for moulds, and when a material has once been 
pasted down the reverse side is not available for study. Fragments 
may be preserved in plastic envelopes, but in order to prevent 
damage by handling they should be tacked to a sheet of cellulose 
acetate before being placed in the envelope. If textiles are too frail 
or brittle to be sewn, they can be exhibited by placing them on a 
sloping surface in a dustproof exhibition case; the surface should be 
faced with rough silk or velvet with the pUe facing upwards so that 
the textile fragments adhere in position. 

It is often desirable to drape tissues for exhibition, and when these 
are frail or tattered they must be given some form of support so that 
they can bear their own weight. It was formerly the practice to use 
a silk net for this purpose, but silk loses its strength in time, and it is 
subject to attack by insects and micro-organisms, whereas materials 
are available today that are not only stronger than silk but Hkely to 
be more permanent. A notable example is Terylene which is un- 
shrinkable and rotproof, and so strong that a very fine quahty can 
be used in repair work. Unfortunately there is some difficulty at the 
present time in applying dye-stuffs to the finished Terylene (though 
it may be dyed during manufacture) and this hmits its use for the 
repair of coloured textiles, such as flags and banners. The traditional 
method of mounting old flags is to sandwich them between two 
pieces of silk net of wide mesh, and when the sewing is finished the 
threads of the net are tinted with water-colour to match the colours 
of the flag. The flag is then displayed with the pole in a horizontal 
position so that the strain on the textiles will be, as far as possible, 
equally distributed. Terylene would be a more suitable material for 

that Diastase should be used in 0*2 per cent, solution in water containing 0-3 per cent, of 
common salt. Stained textiles are soaked for about an hour in lukewarm water (without 
wetting-agent) and then placed in the Diastase-salt solution at pH 6-8-7-I (see Appendix 
V) until the starch or protein is rendered soluble, when the treatment may be completed 
by rinsing and drying or washing with a detergent as in 6 (above). See also Appendix 1 . 3. 



this purpose than silk if the difficulty of tinting the threads could be 
overcome. But this is not the fmal solution of the problem, and the 
method is likely to be superseded. It would seem to be unwise to 
expose a valuable flag where it may be attacked by moth and harbour 
dirt; the most satisfactory solution seems to be to have a copy of the 
original made and hung, and to exhibit the original under glass 
nearby, as has been done so successfully with the relics of the Black 
Prince in Canterbury Cathedral.* 

As a general principle, modem synthetic fibres should be used for 
the repair of old textiles whenever possible, as this makes it easy at 
any time to distinguish between what is old and what is new. 

Teryiene net was used in repairing and strengthening the fifteenth- 
century Pyx Cloth (Pi. 15), a frail textile of drawn-thread work 
which was required not only to support its own weight, but that 
of four wooden balls, one at each comer of the square. The balls were 
tasselled and covered with gesso and gold leaf. A Teryiene net re- 
inforcement was sewn on the reverse side around the central hole 
where the strain was greatest, and some reinforcement was also done 
at the comers; since both the textile and the Teryiene were of a pale 
ivory colour the repair work was invisible at a distance of a few 
feet. All the sewing repair work was done with Teryiene thread. 


Gold threads have been found to vary considerably in structure 
and quality. Sometimes they were made by winding a high-carat gold 
aroimd silk thread (as in the case of the Durham vestments) and it 
was the custom to flatten such threads by hammering in order to 
enrich the appearance and convey the impression of a continuous 
sheet of burnished metal. At other times the gold alloy was much 
debased and became so tarmshed that it will not respond to cleaning. 
Gold was sometimes wound on hair or on parchment, and even gilt 
paper has been used as a winding over a thread core. Gold threads 

‘ Restored by The Tower of London Armouries with the co-operation of The Royal 
School of Needlework. The original jupon has been strengthened with Teryiene net, 
dyed to match the old material by The Fibres Development Dept, of I.C.I. Ltd. 



should always be examined closely before any attempt is made to 
clean them as the poorer quahties would not withstand mechanical 
strain of any kind, and would be ruined by contact with water. 

An imusual form of cleaning and mounting was carried out in the 
case of the gold-embroidered stole, maniple, and girdle, dating from 
the tenth century, that form part of the St. Cuthbert rehcs at Durham 
Cathedral. I The work of cleaning involved removal of a hard buckram 
backing to which the frail old embroideries had been pasted at some 
time in an effort to preserve them. The buckram had shnmk, and 
caused the embroideries to cockle, and they were in a very brittle 
condition possibly due to excess of adhesive or glue from the buck- 
ram. The first step was to reinforce the embroidery by facing it with 
Japanese mulberry tissue. For this purpose it was necessary to choose 
an adhesive that would not be softened by water and that could 
easily be removed at the conclusion of operations without affecting 
the dye-stuffs. Polyvinyl acetate (dissolved in a mixed solvent con- 
sisting of toluene and acetone 95:5) satisfied these requirements, and 
by using this adhesive the thin tissue paper was attached as intimately 
as possible and the surface fixed securely. After the adhesive had dried, 
the embroideries were laid face down on glass, and relaxed under 
sheets of wet blotting-paper imtil the paste had softened to such a 
degree that the buckram could be detached. Residual lumps of softened 
paste that were adhering among the embroidery threads had to be 
picked out with forceps, and the back of the embroidery was then 
cleaned with saponin froth. What remained of the original lining 
was left in situ. The textiles were dried between sheets of warm 
blotting-paper and flattened under glass plates. 

The next stage was to release the mulberry paper by the apphca- 
tion of the mixed solvent, and it was found on removal to bear an 
imprint of the gold embroidery in dust.^ To clean the gold stiU 
further, fresh mulberry paper was apphed with polyvinyl acetate, 
and after one or two apphcations the richness of the gold threads was 

' The Relics of St. Cuthbert, ed. by C. F. Battiscombe, 1956. 

^ The dust contained silver sulphide. This arose from tarnishing of the silver consti- 
tuent of the gold alloy that had been used in making the thread. 



restored to a remarkable degree. The silk threads, however, were 
stiU very dirty. These were cleaned by several apphcations of saponin 
froth, and by this means it was possible to recover the natural sheen 
of the silk and to reveal colours that had been concealed by the grime 
of ages. The embroideries were then sterilized in the thymol chamber 
preparatory to mounting. 

The Durham embroideries are composite in structure; in the main 
length of the stole and maniple there is embroidery on the face only, 
but in each case there are terminal panels with embroidery back and 
front and all the sections are edged with gold braid. For the purpose 
of mounting, the gold braids were removed and the back and front 
of the terminal panels separated. As the original foundation material 
of the embroidery had perished it had to be strengthened with a silk 
gauze backing. A fine silk gauze was stretched in an embroidery 
frame; the main section of the maniple was laid on this face upwards, 
and brown silk retaining threads were passed across at intervals. The 
four sections of the terminal panels were sewn down in the same way, 
separately, and the silk gauze reinforcement was cut all round each 
section, close to the embroidery. The mounting of the maniple was 
carried out as follows: a stronger and more open silk gauze was 
stretched in the embroidery frame and the main section of the maniple 
and the two front terminal panels were sewn on this in their correct 
relative positions. The frame was inverted, and a maroon silk lining 
similar to the original was sewn on the gauze along the back of the 
maniple, and the back sections of the terminal panels added. After 
this the edging braids were replaced (see Pi. 14 a and b). 

The stronger gauze with the mample in position was then stretched 
tightly in a thin sUp frame and this was inserted, for exhibition pur- 
poses, in the channelling of a heavy wooden frame glazed on both 

This manner of mounting was devised by the late Mrs. Guy 
Antrobus. The reinforcement of silk gauze gives adequate support 
to the frail needlework, and the textiles are sealed from the air and 
out of contact with the glass. The Durham embroideries are the 
oldest and finest extant and it would appear that everything possible 



has been done to ensure their preservation. Had the work been done 
today the only improvement that might have been introduced would 
have been to use Terylene gauze and thread in place of the silk. 


No chapter on the conservation of textiles would be complete 
without reference to the damage that is done by exposure to hght in 
causing tissues to become weakened and dye-stuffs to fade. 

AU natural fibres, animal and vegetable alik e, gradually lose their 
strength on exposure to sunlight. This is illustrated in the following 
table pubhshed by Miss Ehzabeth Stromberg,i which, even in the 
absence of experimental details, gives some idea of the rate of decay 
and the relative resistance of the various types of fibre; 

Sunlight Time to reduce Tearing Strength 50 per cent. 

Silk, cultivated . 

, 200 hours 


400 „ 


940 „ 


990 „ 


. 1,200 „ 

Wool, chrome-dyed 

. 1,900 „ 

Actually, the rate of deterioration varies with the intensity of the 
hght and the temperature and humidity of the atmosphere, but it is, 
above all, the invisible ultra-violet constituent of sunhght that is the 
prime cause of the weakening of textile fibres. In the case of dyed 
textiles, it is well known that certain dye-stuffs have a protective 
action in slowing down tendering while others accelerate this 
weakening of the tissue on exposure to hght. 

In the fading of dyed textiles the factors mentioned above are all 
of importance but the nature of the complex formed between the 
dye-stuff and the substrate must also be taken into account. When 
exposed to hght, fading will occur if this complex is capable of 

' Stromberg, Miss Elizabeth, I.C.O.M. News, 1950, 3 . No. 3. See also Harrison, L. S., 
Report on the Deteriorating Effect of Modern Light Sotirces^ Metropolitan Museum of Art, 
New York, 28, N.Y., 1953. 



absorbing radiant energy which brings about a molecular change. 
It should be noted that the spectral absorption of the complex in the 
dyed textile need not necessarily be the same as it is for the dye-stuff 
in isolation. These considerations apply with equal force to any form 
of hghting, natural or artificial, but owing to the relatively low 
intensity of artificial hght sources, the rate of change of colour may 
be so reduced that it can only be assessed after a prolonged period 
of time. Miss Stromberg has drawn attention to the cumulative 
effect upon museum objects of exposure to weak dayhght and to 
artificial hghting, pointing out that, in certain cases, the result of 
long-term exposure may be Just as serious in causing fading as ex- 
posure to sunhght for a correspondingly shorter time. However, in 
making comparisons of fading, it must be remembered that the 
relative humidity has been shown to have an important bearing on 
the fastness of dye-stuffs to hght. 

A detailed examination of the tendering and fading of textiles 
when exposed to narrow isolated bands of radiation in the extreme 
visible and near ultra-violet is being carried out at the present time 
under the auspices of the International Institute for the Conservation 
of Museum Objects. In a prehminary examination of tungsten and 
fluorescent sources, filtered and unfiltered, comparative tests have 
given an assurance^ that dye-stuffs suffer no abnormal fading when 
exposed to fluorescent hghting. For the internal hghting of cases, 
however, fluorescent tubes are to be preferred to tungsten hghting as 
they are cooler in operation, but it is essential that auxfliary equip- 
ment (chokes, &c.) should be mounted outside the case where, inci- 
dentaUy, it will be readily accessible for servicing. - 

While there is httle to choose between tungsten and fluorescent 
hghting from the point of view of tendering and fading, the intensity 
of the hght source is an important factor. Some suggestions are now 
discussed to indicate how textile collections may be protected from 
the deteriorating action of hght, both natural and artificial. 

' Cooper, B. S., Museums Journal, 1954, 53 » P- 279; G.B.C. Journal, 1956, 23, p. 192. 

^ For practical advice on museum lighting see Use of Fluorescent Lighting in Museums 
(1953). I.C.O.M., UNESCO House, 19, Avenue Kleber, Paris (16^). 

1 12 


when the museum or gallery is closed, collections should be kept 
in darknesss as should all materials in storage. Permanent exhihition 
of valuable textiles is inadvisable unless special precautions are taken 
to reduce the period of exposure to hght as much as possible; tem- 
porary exhibitions are to be preferred. Sunhght should be prevented 
from falling direcdy upon valuable textiles by the use of opaque 
blinds or curtains. The intensity of hghting, natural or artificial, 
should not he excessive. Roof hghting should be controlled by 
louvres or adjustable blinds, or it may be desirable in some cases to 
cut down the intensity of the hghting by colour-washing the glass 
in summer. Where there are vertical windows the hghting may be 
diffused by introducing blinds of the Venetian type, having a system 
of translucent opal plastic slats in place of the traditional wooden 
members; these are hght in appearance and easy to adjust and clean. 
In certain cases protection may be afforded by fitting windows with 
special glass' that has the property of excluding the rays that are most 
harmful. Such glasses cut out a portion of the spectrum and so 
must restrict the overall hghting to some extent, and they are neces- 
sarily shghtly tinted, but they are of value for roof hghting and for 
hghts in churches where, for example, it is desirable to prevent strong 
hght from falling upon tapestries or altar cloths. Special glasses 
materially reduce the tendency to fading, but it should be added that 
no glass is available that will altogether prevent the fading of dye- 


Sulphur dioxide is an impurity in the atmosphere resulting from 
the combustion of fuel, and it readUy forms sulphurous acid which in 
turn is oxidized to sulphuric acid. Mention has already been made of 
its action in causing the deterioration of organic materials such as 
leather and paper, and it is the most potent cause of the tendering of 

' Pilkington Bros. Ltd., St. Helens, Lancs., make several such glasses. The most appro- 
priate for window glazing is called Antisun, which at a thickness of 3/16 inch has a hght 
transmission of 75-80 per cent., greatest in the middle of the spectrum where vision is 
best. This glass absorbs both heat and ultra-violet rays. 



textiles. Cotton seems to be particularly prone to decay from the 
action of sulphur dioxide. Textile collections in cities or near indus- 
trial undertakings are likely to be best conserved when protected 
under glass. When the small concentration of sulphur dioxide within 
a case has been absorbed, no further contamination will take place 
until the case is opened and a fresh supply of polluted air is admitted. 

Iron catalyses the change of sulphur dioxide to sulphuric acid, 
and it is a common observation that where iron nails or tacks have 
been used to fix textiles, the material has rotted more intensively 
where it is next to the iron, as, for example, around the tacks used to 
attach a canvas painting to its stretcher. For this purpose copper tacks 
should be used. But ordinary dust usually contains an appreciable 
quantity of iron in one form or another, and this is an additional 
reason for adopting some form of protection for materials that are 
exhibited in galleries in which the air is not fuUy conditioned. 


In common with other organic materials, textiles are susceptible 
to attacks by moulds, but if the textiles are clean and conditions are 
reasonably dry, there is not likely to be any trouble in this connexion. 
In the event of mould growth being discovered, it can usually be 
arrested by airing the textiles, and any surface growths can be 
removed with a soft brush. If the outbreak is extensive, the textiles 
should be sterihzed with thymol vapour (p. 56) but sterihzation may 
be omitted if the textiles can be washed by an immersion process. If 
textiles are damaged by water, as might arise from a burst water-pipe 
or leaking roof, fungoid attack is hkely to ensue, and the most eflec- 
tive first-aid measure is to expose them in a current of warm dry air. 
For emergencies of this kind the most suitable equipment is the 
electric hair-drier. The use of a heater without ventilation must be 
avoided as it would be likely to encourage and intensify mould 

Insect pests are a more frequent cause of damage to textiles than 
fungoid attack. A nim al fibres, wool and silk, provide foodstuffs for 

B 6157 



1 14 

several varieties of insects, and dry conditions are no serious deterrent 
to their activities. Linens and cottons are normally immune from 
attack, and they should be segregated from the wools and silks to 
facihtate inspection of the more vulnerable materials. But textiles of 
all kinds are hable to be damaged if they should happen to be con- 
taminated with substances that are attractive to the pests, and it is for 
this reason, as much as for any other, that garments are washed or 
cleaned before incorporation in the collection. Textiles should be 
inspected at regular intervals according to a routine, involving im- 
folding, dusting, airing, and sunning, and specimens should be scru- 
tinized to discover any traces of insect attack — wholes in the material, 
presence of grubs, cocoons, or loose silken threads. 

There are three general ways of protecting textiles from insects. 
The first is to isolate the textile — to parcel it up so that insects cannot 
get at it. The second is to use insecticides, and the third is to use an 
inhibitor that is either so distasteful to the insect or so toxic that pro- 
tection is assured. This last method involves the use of a proofing 
reagent that can be permanently fixed on the textile fibre. This 
necessitates the use of special apparatus and is a technique that can 
only be apphed to new material in the course of manufacture. 

Isolation of the textile. This is only a practicable proposition in the 
case of individual specimens, and, even then, there is always the 
possibihty of infection being introduced in the parcelling and of 
activity proceeding unrecognized. It is advisable to include a quantity 
of volatile insecticide in the parcel. For wrapping textiles, paper is 
as effective as transparent plastic, but the latter is stronger, and has 
the added advantage that the contents are more readily inspected. 

When Polythene bags are used, it is most important that the tex- 
tiles are quite dry, as moisture cannot escape through Polythene, and 
if the textiles are damp when packed, they are almost certain to be- 
come mildewed. The proofed bags sold for the storage of furs are 
also satisfactory for textiles, but it is as well to put about a cupful of 
dichlorobenzene crystals in the bottom of the bag to ensure against 
the risk of moth eggs being present. Protection by sandwiching 
material passe-partout between glasses provides an alternative method 



of isolation, and insecticides are normally not required here as the 
textile is exposed continuously to view. 

Use of insecticides. Apart from vacuum fumigation that can be ap- 
phed on a large scale to protect costumes or ethnographical material 
in bulk (see Wood, p. 123), the most effective protection is given 
by the presence of a volatile insecticide such as dichl orobenzene, 
which, though perhaps a trifle too volatile, is otherwi^~ satisfactory ^ 
and efficient. It can be filled into cotton bags and hung in wardrobes, 
or it may be scattered in drawers and chests between layers of paper. 
This substance is very effective against the chief insect pest, the 
clothes moth, and is also effective against dermestid beede. It has, 
however, a rather overpowering smell which causes headaches, and 
as it is a hver poison it is important to avoid working for long over 
an open case in which this chemical is used.^ 

For materials that cannot be protected in this way the alternative 
is routine spraying using, preferably, a power spray with a fish-tail 
nozzle, and working at a pressure to give fine atomization. The gun 
is held at an angle of about 45° to the textile, so that a cloud of insecti- 
cide floats on to it, rather than a direct blast which might be damag- 
ing. Many different insecricides are available for this purpose, the most 
satisfactory being those made up in a colourless and odourless petro- 
leum distillate. Pyrethrum extracts have an immediate action upon 
moths, DDT is slower to take effect, and insectides of a mixed type 
containing both ingredients are probably best (see p. 29). Very httle 
of the spray material condenses on the textile, and there is no evidence 
that the regular use of the spray results in any staining or change in 
colour. Insecticidal dusts are Hable to contaminate the material un- 
necessarily and are not recommended. 

• In Moth-proofing from the Entomological Point of View, Vezelinstituut T.N.O., Delft, 

1951, Dr. A. D. J. Meeuse has shown that even fumigation with hydrogen qranide is no 
guarantee against survival of moth eggs, and that camphor, naphthalene, and dichloro- 
benzene act on moths as respiratory poisons when used in high concentration, but have 
no repellent action in low concentration, though there are indications that dichloroben- 
zene is more effective than the other two. In ‘Insect Pests of Books and Paper’, Archives, 

1952, 7, p. 19, A. W. McKenny Hughes recommends using dichlorobenzene crystals in 
an air-tight cupboard at a concentration of i lb. to 10 cub. ft. of cupboard space. 



Wood has been exploited by man since palaeolithic times. Man has 
done almost everything with wood — made his home of it, made 
fires with it to cook his meals, used it for utensils, weapons, and 
implements, built ships and bridges of it, used it for making 
vehicles, furniture, art objects, and musical instruments, and, in 
modem times, transformed it into paper and even clothing. It is not 
surprising, therefore, that wood forms a large part of the collections 
in museums — particularly in ethnographical and folk museums — and 
its conservation is a matter of considerable importance. 

Being of organic origin, wood normally decays under combined 
biological and chemical attack when buried in the ground, but in 
exceptional circumstances it has been found to survive prolonged 
exposure to extremes of dryness or wemess. Worked timbers exist 
today in the dry tombs of Egypt that date back to the early dynasties, 
and it is somewhat of a surprise to find that these are still sound and 
often so fresh in appearance that they might be taken for timbers of 
modem origin. Timber that has been buried in wet peat bogs for 
a long period of time may retain its general shape and size. The 
absence of air inhibits fungoid attack, but profound changes may 
have taken place in chemical composition and microstructure, result- 
ing in a loss of strength. Such timber can be dried, however, and its 
original appearance restored by laboratory treatment. In sharp con- 
trast, wood completely decays when buried in damp sand that is well 
aerated. In this case decay is due in the main to fungoid attack. The 
excavation of the seventh-century ship burial at Sutton Hoo yielded 
many notable treasures (now in the British Museum) but the ship 
itself had succumbed, leaving only rusty iron nails and a stained 
impression in the sand to mark where the wooden planks had rested. 



No such extreme changes are to be expected in wood kept indoors. 
Under such circumstances the commonest form of deterioration, in 
temperate climates at all events, is due to the dimensional changes 
accompanying variations in the relative humidity of the atmosphere. 
These restdt in warping and sphtting, features that are related to the 
hygroscopic nature of the material, and to the fact that it has a grained 


In order to study the conservation of wooden objects it is necessary 
to be famihar with certain characteristics of the raw material. Wood 
is an anisotropic substance, exhibiting different degrees of hard- 
ness, toughness, &c., in different directions. It has an organized cel- 
lular structure, and as the fibres are for the most part orientated in 
the same direction, grain is a distinctive feature. This varies in ap- 
pearance according to the species of tree and the cut of the timber.^ 
In a longitudinal median plank, two main zones are to be dis- 
tinguished, namely heart-wood and sap-wood, the latter being 
situated towards the edges adjacent to the bark. In practice the direc- 
tion of the cut is not necessarily chosen as being the most economical 
nor as giving the strongest timber, though both these considerations 
may apply; it is often selected in order to give some effect of texture 
or grain for ornamental purposes. 


The saprwood of a tree has a higher moisture content than the 
heart-wood, and when the timber is sawn into planks, the distribu- 
tion of moisture in the planks varies accordingly. If left to dry out 
naturally, the moist parts will shrink more than the drier parts and 
the planks will warp. It is for this reason that they have to be sub- 
mitted to a process of controlled drying known as seasoning. This 
consists in allowing the timber to dry out slowly under conditions 

■ InMiwewm, 1955,8, 3, p. 139, the relationship between cut and grain is illustrated in an 
admirable series of drawings. 



in which distortion is prevented. But wood never loses all its mois- 
ture; the cell moisture^ remains, and, when seasoning has been com- 
pleted, the moisture content of timber is found to vary with the 
relative humidity of the atmosphere surrounding it. The equihbrium 
state of air-dry timber that has been seasoned under the most favour- 
able conditions varies in different regions. Desch^ quotes a moisture 
content of 12-15 per cent, as typical of timber seasoned in the United 
Kingdom and most regions of the United States, whereas in the 
more humid tropics, e.g. Malaya, 14-18 per cent, is characteristic, and 
in hot arid regions the moisture content may be as low as 8-12 
per cent. 

In the early stages of seasoning sap-wood is particularly vulnerable 
to attack by fungi and insects, not only because of its higher moisture 
content, but because it contains a greater concentration of nutrients. 
Kiln-seasoning is general today instead of air-drying, and as it speeds 
up the process there is less chance of infection; it has the additional 
advantage that the residual moisture content of the wood can be 
adjusted to suit the particular purpose for which the wood is required 
and the climate in which it is to be used. 


The most important factor in preserving seasoned timber is to 
avoid exposing it to wide variations in atmospheric humidity. Under 
such conditions it alternately absorbs and gives up moisture. While 
this causes httle dimensional change along the grain, it causes expan- 
sion and contraction across the grain with consequent warping. If 
one side of a plank is protected from the atmosphere, the other side 
will absorb or give up moisture more quickly than the protected 
side and warping will be intensified. Humid conditions will cause 
expansion across the grain on the exposed side, so that the protected 
side becomes concave. Conversely, where conditions are very dry 
the protected side will become convex. This explains the ‘cupping’ 

' i.e. the moisture contained within the cells as distinct from the moisture held between 
the cells. 

* Desch, H. E., Timber, Its Structure and Properties, 1938, p. 94. 


1 19 

that takes place in the constituent planks of a wooden panel painted 
only on one side, when exposed in an atmosphere that is too dry. 
Varying relative humidity tends to cause back and forward move- 
ments which are a significant factor in the hfe of panel p ainting s 
(p. 4). If a panel is restrained in a frame by glue or nails, changing 
humidity may cause the wood to crack or become distorted. 

The elimination of warping is a lengthy and uncertain process,* 
although at times it has to be attempted in the interests of conserva- 
tion. The operation may take several weeks or even months. The 
treatment consists in moistening the concave side so that water is 
slowly absorbed; this swells the tissue and encourages the wood to 
return to the flat condition, and it is then kept flat under weights 
until dry. But even if this operation appears at first to be successful 
the tendency will be for the wood to return gradually to the warped 
condition unless some form of cradling can be used to confine the 
stresses to one plane — see pp. 161-2. 

A simple form of metal reinforcement that is often quite effective 
and may be applied where the wood is of reasonable thickness is to 
screw a series of angle irons to the back of the panel at right angles to 
the grain by countersunk screws passing through slots (not holes) in 
the metal. The edges of the slots should be slightly chamfered. This 
arrangement admits of lateral movements of the wood and prevents 
the panel from cupping. The metal and screws must be of such a 
nature as not to loose freedom of movement by corrosion or otherwise. 

When, as in certain types of furniture, the wood is painted all over 
or French-pohshed, all the surfaces are equally protected and move- 
ment of the constituent timbers is reduced to a minimum. In the case 
of synthetic grainless boards that have no directional properties, or 
plywood having an adequate number of layers, a balance of forces 
is achieved structurally, and warping is prevented. Large thin boards 
may buckle imder their own weight, but this is purely a mechanical 
matter which can be dealt with by arranging for suitable supports. 

’ When wood is old and has been in a warped condition for a long time, a modifica- 
tion of cell structure may have occurred which precludes the possibihty of a successful 
restoration. Wood in this condition is said to have acquired a ‘permanent set’. 




A certain degree of dampness is required before fungi can grow on 
materials supplying the necessary nutrients. This has already been 
referred to in the case of paper and leather, and it is equally true of 
wood. Structural timber is most prone to attack because it may be- 
come wet owing to the inadequate provision of damp-proof courses, 
and large wooden objects, such as canoes, totem poles, and carvings 
exposed out of doors, are much more likely to be attacked by micro- 
fungi than objects kept indoors. The fungi attacking such timbers 
may be of the virulent ‘dry rot’ group that is so difficult to eradicate 
without wholesale sacrifice of the infected material. Preventive 
measures take the form of isolating the wood, if possible, by the 
introduction of a damp-proof course of slate, lead, &c., and by pro- 
tecting any end-wood by impregnating it with a waterproofing agent 
such as linseed oil, wax, or a water-repellent sihcone varnish. ^ When 
timbers are exposed in the open, advantage should be taken of any 
possibihty of deflecting water by dripstones or courses. When wood 
is in contact with the ground, it should be impregnated with a fungi- 
cide, and any part under ground level encased in waterproof cement. 

The following fungicides are colourless and odourless and will not 
stain wood, brickwork, or plaster. 

(A) 3-6 oz. of commercial sodium fluoride dissolved in one gallon 
of water and appHed cold by brush (two coats), or 

(B) Half a pound of commercial acid magnesium fluoride dis- 
solved in one gallon of water.^ 

For details of these processes and of other methods of dealing with 
an outbreak of dry rot, the reader is advised to consult the pubhca- 
tions of the Forest Products Research Laboratory.^ No satisfactory 
account of methods of conserving timber would be possible without 
frequent reference to the valuable pubhcations of this laboratory, 

’ Midland Silicones, Ltd., 19 Upper Brook Street, London, W.i. 

^ A wooden tub must be used in making this solution as it attacks metal. 

^ Forest Products Research Laboratory, Dry Rot in Buildings, Recognition, Prevention 
and Cure, Leaflet No. 6, revised April I947J Dty Rot in Wood, Research Bulletin No. i, 
4th edition, 1945. 



which are based upon wide experience and research, and cover all 
aspects of the subject. 

When moulds are growing on art objects in the museum, ethno- 
graphical specimens, &c., it will usually be found that they are feeding 
upon oil, grease, or the binding medium of gesso, rather than on the 
wood itself, and growth may be arrested by cleaning and applying 
fungicide. But the treatment should not stop here. It is equally im- 
portant to take measures to prevent a recurrence of the outbreak 
by improving ventilation. 


Insect pests of timber are a greater menace to museum objects than 
moulds, because the possibiHty ofinfection is ever present. The nature 
of the damage will depend on the type of infestation, which varies in 
different parts of the world. The commonest pests in Great Britain 
are various types of wood-boring beetle, and the damage is done by 
their grubs which tunnel into the wood, converting sound tissue into 
powder. The grubs are popularly known as ‘wood-worm’. 

If the attack is not checked at an early stage, the wood may fall to 
pieces and may be found to be riddled with insect channels filled with 
powder and frass resembling fine sawdust. It is usually the presence 
of powder of a hght colour that provides the first indication of attack; 
if structural timber is inaccessible, the first indication may be that 
a piece of timber gives way and falls to the ground, as happened 
in the case of a wooden rib from the roof of the North Transept of 
York Minster in 1934. When this occurs the attack will have already 
proceeded to great lengths, and it can only be dealt with by an expen- 
sive operation involving wholesale impregnation of the old sound 
wood with insecticides, and by making extensive replacements of 
the infected wood. 

Furniture should be inspected regularly to ensure that it is free 
from ‘worm’, paying special attention to plywood back-boards and 
also to any sap-wood that may be present. It is not uncommon to 
find the first evidence of a wood-beetle infection in the ply-backing 


of a framed picture, or in the plywood that has been used to repair 
the back of a chest or a bureau. When discovered, infection should be 
dealt with at once, either by burning the plywood or by treating it 
with a rehable insecticide (see p. 125). 

The discovery of worm holes in wood is not necessarily an indica- 
tion that insects are still active. Sometimes they die off before all the 
wood has been destroyed, but it is never an easy matter to be certain 
that activity has ceased. Clean wood dust falling firom an old hole 
may suggest activity which, in fact, does not exist; but if an object 
that has been lying undisturbed -with accumulations of dirty wood 
powder is seen to have some clean wood dust upon it, this will be a 
clear indication that insects are still ahve in the wood. In all cases 
where there is any doubt, it is safer to assume the worst and the object 
should be hfted carefully into the open air, brushed clean, and treated 
with an insecticide. 

Before it can be taken for granted that sterilization has been com- 
pletely effective and that no viable eggs remain, the wood must be 
kept under regular observation during at least one complete hfe-cycle 
of the insect, but as insects develop at different rates this involves 
identification. For example, the common powder-post beetle [Lyctus) 
breeds through from egg to adult in three, months to a year, whereas 
in the case of the death-watch beede [Xestobium ) — the major pest 
of inaccessible structural timbers — development is much slower, 
taking possibly two years or more to complete. The Ufe cycle of the 
common furniture^ beetle {Anohium) may also run to two years or 
more. Wood-boring insects may be identified by the size of the holes 
they leave on emerging to lay eggs, or by the materials attacked, or 
by the nature and extent of the outbreak, age of the timber, &c., but 
the most certain method is a microscopic study of the grub or full- 
grown insect and comparison vtith photomicrographs and drawings 
in the standard works of reference. * 

As an aid to long-term inspection and sterilization all surface 

* Fisher, R. C., ‘Household Timber Insects, Recognition, Prevention, and Control of 
Infestation’, The Sanitarian, November 1949; see also the leaflets of the Forest Products 
Research Laboratory devoted to specific pests. 



powder and the frass should be removed and the old holes filled with 
soft wax (see p. 125) so that any fresh ones will be at once apparent. 
In adding to a museum collection, the greatest care must be taken to 
exclude material that might introduce infection imtil such time as it 
has been sterilized. When wood is found to be attacked, treatment 
should be carried out as soon as possible; the longer the delay, the 
more difficult it will be to bring the infestation under control. 


Insects may be killed by rise^f temperature, by reduction of pres- 
sure (vacuum treatment), and by poisoning with gas (fumigation), 
or hquids (spraying; impregnation). It may be stated in general that 
fumigation methods are immediately effective but confer no lasting 
protection, whereas impregnation methods may be slower to take 
effect, but the poison remains and confers protection over a period. 

Sterilization by fumigation. Sterihzation may be carried out in a 
sealed gas chamber, and with suitable equipment it is possible to 
reduce the pressure to a partial vacuum before admitting the gas, 
which ensures that the fumigant will penetrate in a reasonable time 
into the objects under treatment. 

Hydrogen cyanide gas is used in such a chamber at the British 
Museum (Pi. 16) for dealing with ethnographical material in bulk, 
mainly textiles, skins, feathers, and small wooden objects, for which 
a twenty-four hours’ exposure has been found adequate. With denser 
materials such as heavy timber, methyl bromide would be likely to 
give better penetration, but methyl bromide has the disadvantage 
that, in dealing with ethnographical specimens which may contain 
feathers, leather, &c., or with upholstery containing horse hair, it tends 
to form compounds having a very impleasant odour. This rules it 
out for museum work where objects in storage have to be frequently 
examined. When large objects have to be sterilized by fumigation 
and no special plant exists for the purpose, it may be possible to 
arrange for this to be carried out professionally by a firm specializing 
in such work.* 

’ London Fumigation Co., Marlowe House .Lloyd’s Avenue, E.C. 3. 



Carbon disulphide is a reliable insecticide that can be used for the 
fumigation of valuable objects of all kinds as it does not harm the 
most dehcate material. It is a clear Hquid at normal temperatures, 
having a lotv vapour pressure (B.P. 46° C.) and an unpleasant odour. 
As it evaporates readily to form explosive mixtures with air,* rigid 
precautions must be taken to exclude naked hghts and prevent 
smoking in any room in which this fumigant is used. Some museums 
are equipped with special vacuum plant for fumigating with carbon 
disulphide vapour. The special plant is enclosed and therefore 
safe and assures good penetration. It is possible to fumigate with 
this insecticide at normal temperature and pressure with impro- 
vised equipment. This type of fumigation is conducted as follows. 
Dishes of the hquid are exposed above worm-infested wood in an 
air-tight cupboard or box which is then sealed up, the doors being 
pasted round with adhesive paper or Polythene tape. A Polythene bag 
may be used instead of a cupboard provided it can be adequately 
sealed with tape. Liquid carbon disulphide should be used in the 
proportion of i ounce to about 8 cub. ft. of air space in the chamber. 
The hquid volathizes in the course of a few hours, and the heavy 
vapour pours down over the wood and tends to accumulate near the 
base of the cupboard. The time of exposure required for fumigation 
is from 14 days to three weeks, and a fresh amount of Hquid carbon 
disulphide should be added at the end of the first week to take the 
place of that which has evaporated. After fumigation the unpleasant 
smell disappears on exposing the wood to air for a short time. 

It will be obvious that fumigation by carbon disulphide is particu- 
larly effective for sterihzing boards that can be treated in a horizontal 
position in a shallow box, and as this fumigant is without action on 
paint or varnish the method is very useful for the treatment of worm- 
eaten panel paintings. 

Liquid carbon disulphide should be stored in a cool place and 
protected from dayhght as this causes the Hquid to develop acidity. 

For the steriHzation of wooden objects that are not decorated with 

* Carbon disulphide is a dangerous chemical and fumigations involving its use are best 
left in the hands of trained personnel. 



fugitive colours, a non-inflammable solution of one part of carbon 
disiflphide in four parts of carbon tetrachloride may be recom- 
mended. While this is not quite so toxic to insects as pure carbon 
disulphide, equally satisfactory results are obtainable either by using 
a greater concentration or extending the period of exposure. 

Sterilization by impregnation. Impregnation with UquidLinsecticides 
may be carried out by using a pipette or a syringe to inject the hquid 
into the holes. Alternatively, the insecticide may be brushed in. 
Large imdecorated timbers may require to be drilled in inconspicuous 
parts to enable the insecticide to penetrate. 

Suitable hquid insecticides are those containing DDT, BHC (gam- 
mexane) , pentachlorophenol and its derivatives, chloronaphthalenes, 
or the metalhc naphthenates. Many rehable mixtures are available 
commercially. ^ If apphed thoroughly to all insect holes, the impregna- 
tion method may be as effective as fumigation, but before using an 
insecticide it should be tested to make sure that it is non-staining and 
that it does not soften paint or damage any decorations on the wood 
with which it might come in contact. 

For fflhng the holes after treatment, the best material is usually a 
soft wax to which an insecticide has been added. A suitable prepara- 
tion for fflhng the holes can be made by stirring DDT powder 
into molten beeswax, bleached or appropriately coloured to match 
the timber. When soUdifled, the wax paste may easily be spread into 
the worm holes with a blunt knife. If the wood requires strengthening 
(see below) this ought to be attended to prior to waxing. 

Sterilization by Spraying. While spraying is not very effective for 
kflhng pests that have already bored their way into wood, there is 
evidence that spraying with a 2 per cent, aqueous emulsion of DDT 
can prevent the spread of an outbreak of Lyctus.- 

Proofing of wood against insect pests. Proofing is essential in the case 
of timber that may be exposed to attack by termites (‘white ants’) — 
the most injurious wood-feeding insects known. Creosote conform- 

‘ See Wood Preservations Register, ipMh]is]:iedhY Timber and Plywood, 194-200 Bishops- 
gate, London, E.C.2. 

^ Forest Products Research Laboratory, Leaflet No. 43, 1948. 



ing to a standard specification* is used for this purpose but to be 
entirely effective the wood should be impregnated under pressure, 
and the creosote retained by the timber in certain minimum amounts. 
MetaUic naphthenates are also used for pressure impregnation. For 
the protection of museum objects of wood such as furniture that can- 
not be impregnated under pressure, the only alternative is to saturate 
the wood as completely as possible by dipping or brushing, using a 
non-staining insecticide that will penetrate deeply into the timber. 
Protection is hmited in proportion to the depth of penetration of 
the preservative. Such an insecticidal mixture may be prepared by 
dissolving pentachlorophenol in a petroleum distdlate. This leaves 
the wood clean after treatment, but it is emphasized (F.P.R. Leaflet 3 8) 
that a certain minimum period of immersion is necessary if this 
treatment is to be entirely effective. For thin wood where glue is 
absent a hot 4 per cent, aqueous solution of zinc chloride or sodium 
fluoride may be used, the object being dipped into the solution; but 
with aqueous solutions penetration is less effective and the degree of 
protection usually less satisfactory than that obtainable by the use 
of petroleum solutions or chloronaphthenates. 


Wood that has been weakened by the attack of fimgi or insect 
pests may be strengthened by impregnation with a consohdating 
agent or it may be reinforced mechanically. The nature of the speci- 
men and the condition of the timber will determine which of the two 
methods is likely to be the more satisfactory, but in many cases it 
will be found that both mechamcal strengthening and consohdation 
are required in order to restore sohdity. When this is the case 
impregnation may sometimes precede reinforcement, but sometimes 
it will be more appropriately apphed after the wood has been 
mechanically strengthened. 

' British Standard Specification Number 144. See also Forest Products Research 
Laboratory Leaflet No. 3 8 which gives an account of the habits and habitat of termites. 
References are given in this leaflet to the problem as it presents itself in Austraha, India, 
Malaya, West Indies, and U.S.A., as well as in Great Britain. 



Consolidation by mechanical reinforcement 

Of the mechanical methods for strengthening timber, the follow- 
ing are the most important.^ 

1. Do welling either with metal or wooden pegs, and refacing if 
necessary with wood. 

2. Inlaying across cracks with soHd X-shaped wedges to prevent 
the cracks from opening; or covering the joints by ‘buttons’ of wood, 
i.e. small palettes ca. 3 in. by 2 in. glued across the cracks. 

3. Reinforcing with wooden splints, glued and/or screwed to the 
old wood. Special bracket irons, angle irons, &c., may be useful in 
repairing furniture. 

4. Stopping irregular cavities with a gap-filling cement (see 
repairs, p. 132). 

Consolidation by impregnation 

Methods of consoHdation by impregnation may be applied to all 
porous materials and are particularly suitable for dealing with intri- 
cate shapes, carvings that have suffered erosion by exposure, or wood 
that is riddled with worm holes. Impregnation is carried out by 
immersion, injection, or the solutions may be appHed with a brush. 

Where the surface is covered with loose gesso and gold leaf it can 
only be satisfactorily consohdated by impregnation with a substance 
that penetrates easily, and synthetic consoHdating agents dissolved in 
organic solvents are recommended in this case. Painted gesso is in the 
same category, but the paint must be tested to ensure that it will not 
be softened or dissolved by the solvent. If there is any sign of its being 
attacked it may be possible to change the solvent for one less active, 
or waxing may be the preferable technique. 

Impregnation by wax. When mechanical repairs are necessary they 
are always carried out before the wood is waxed. For many purposes 
the wax-bath is a most effective form of treatment, but for large 
objects special equipment is required. The bath may be a capacious. 

* This matter, in so far as it concerns painted panels, is given detailed attention in 
Museum, 1955, 8, 3, pp. 139 seq. 



rectangular iron tank, heated electrically and thermostatically con- 
trolled so that the temperature of the wax is not allowed to rise above 
predetermined hmits. There should also be an independent safety 
device to switch off the electric supply in case of failure on the part 
of the thermostat. Block and tackle will be required for raising and 
lowering heavy objects, and with this equipment it is possible to deal 
conveniently with hfe-sized wooden sculptures. 

The composition of the wax mixture may be varied but the main 
ingredient is usually unbleached beeswax to which has been added 
a resin in quantities not exceeding 50 per cent. 

The object to be impregnated, which must of course be quite dry, 
is lowered into the molten wax and kept beneath the surface by 
weights or otherwise. As the temperature gradually rises the air is 
expelled in a stream of bubbles and wax enters the pores of the 
wood. Any traces of moisture are driven off by maintaining the 
temperature at about 105° C. till bubbhng ceases, and it may then 
be allowed to rise to about 120° C. The time of immersion de- 
pends on the porosity and bulk of the timber. When the im- 
pregnation is considered to be complete the heating is shut off, 
the object raised from the tank, and the hot wax allowed to drain 
away. Finally, surplus wax is removed from the surface with tur- 

Waxing by immersion has been developed on a large scale in 
Baltimore by Rosen. ^ The process is also apphed with much success 
in the Central Laboratory of the Belgian Museums, Brussels,^ as a 
routine method of treating church furniture, architectural material, 
wooden carvings, and the like (Pi. 17). 

For the impregnation of small objects the wax immersion process 
can be carried out using improvised equipment, but a note of warn- 
ing is necessary about a potential source of danger when the tank is 
heated from below. If moisture has been introduced it will accumulate 

’ Rosen, D., The Preservation of Wood Sculpture: The Wax Immersion Method’, 
Journal of the Walters Art Gallery, 1950-1, pp. 4.5-71. 

^ Private commvmication from Dr. Paul Coremans, Director of the Central Laboratory 
of the Belgian Museums, Brussels. 

I 7- V A 1 r()\i 1 M p u i,(, N A'l I N(; i.A !u; h woo 1) I N M (; u R 1 s w n ii mo i/n,N wax ( i a ho k a i o i r i: 

MUSU,S 1)1 lil-.H.lQUl,) 


Adjacenr sections of ash originally of the same diameter. Lc/n Untreated, showing 
considerable shrinkage. Right: Pre-treated to prevent shrinkage on drying 



at the bottom of the tank as the wax sohdifies; on reheating, 
steam pressure may be generated beneath the semi-sohd wax and 
may cause an explosion. This danger can be mitigated by placing 
a metal rod upright in the wax as it cools; the rod is withdrawn and 
most of the water decanted away before the wax is reheated. It must 
also be remembered that the molten wax mixture is inflammable, 
and that serious bums can be caused by splashing. When wax is being 
melted over an open flame, an asbestos board should be at hand to 
place over the vessel in the event of the wax catching fire. 

Wax-resin mixtures are inert and very stable, and being water- 
proof they afford protection against damp. They are thus of value in 
preventing movements in wood due to changes in the relative 
humidity of the atmosphere. Waxing has certain disadvantages, how- 
ever, unless the objects are kept in cool surroundings. Even if the 
surface of a waxed object has been cleaned after immersion in the 
bath so that only the thinnest film of wax remains, there is a tendency 
for wax to creep out gradually from the inside, and in a matter of 
time the surface will become sticky and tend to collect dirt. In 
addition, if a waxed object is exposed to the sun or exhibited in a glass 
case that is ht internally, the specimen may develop an objectionable 
gloss with rise of temperature. A further disadvantage of wax mix- 
tures is the fact that they have a high refractive index and lower the 
tone of colours. The optical qualities of hght colours are best con- 
served by using materials having a lower refractive index than wax, 
and in certain cases colourless synthetic resins may be preferred. 

Impregnation by synthetic materials. Thin varnishes are used as impreg- 
nating agents to consohdate wood either before or after mechanical 
repairs have been carried out. 

By making a test with coloured ink it can be shown that wood 
absorbs hquids much more quickly in the direction along the grain, 
and this fact is borne in mind when impregnation is carried out by 
brush or by injection. The thin hquid is fed into the end-wood so 
that it permeates the timber as far as possible, and the work is com- 
pleted by coating the surfaces; on volatflization of the solvent the 
resin remains to consohdate the tissue. The following colourless 





synthetic resin varnishes are suitable consolidants for wood — poly- 
vinyl acetate dissolved in a mixed solvent of 9 vols. of toluene and 
I of acetone, or Bedacryl 122 diluted with toluene to a suitable 
consistency. Other materials that have become available recently for 
the impregnation of dry wood are the polyester resins such as Marco 
S.B. 26C2 or Bakehte 17449.3 These are supphed as mobile liquids 
which set in the pores of the wood to hard insoluble sohds, as a result 
of a catalysed chemical reaction. One of these resins was used with 
success by Dr. G. A. Horridge for the consolidation of the dehgrnfied 
wood of Cody’s Tree at the Royal Aircraft Estabhshment, Fam- 
borough.4 An object that has been tunnelled by wood-worm may be 
strengthened by running the solution into the worm holes with a 
fountain-pen filler or drawn-out glass tube as is done in impregnating 
with insecticides. 

In order to strengthen dense timber the impregnation should be 
conducted in an apparatus from which the air has been withdrawn. 
For this purpose a metal vacuum tank is used. If the tank is much 
larger that the object, a vessel of smaller dimensions is used as the 
container in order to economize in the use of solvent, and this vessel 
is placed inside the larger tank which is then evacuated. When the 
pressure has been reduced to the required degree, as recorded by the 
pressure gauge, the impregnating solution is admitted gradually 
through a tube to the smaller vessel until the object is completely 
covered with the solution; it is allowed to remain there for about 
an hour, so that the solution is absorbed as far as possible by the wood. 
Air is then admitted and this drives the consolidant farther into the 
wood, ^^hen the tank reaches atmospheric pressure again, the object 
is hfted out of the container and allowed to drain; it can be dried 
without acquinng a gloss by leaving it in an atmosphere of the 
solvent. This process, which is decidedly the rnost effective method 

' Imperial Chemical Industries Ltd. Supplied as a 40 per cent, solution in xylene which 
is too viscous for most purposes and requires dilution. 

^ Messrs. Scott Bader & Co. Ltd., 109 Kingsway, W.C. 2. 

3 Messrs. BakeUte Co. Ltd., 12 Grosvenor Gardens, S.W. i. 

* Royal Aircraft Estabhshment, Famborough. Technical Note 1236, 1954. 



of Strengthening wood, is of general application to all kinds of porous 
material whether of organic or inorganic origin. 

Apparatus for vacuum impregnation can be set up on a small scale 
using ordinary laboratory equipment. AH that is required is a vacuum 
desiccator provided with a two-way tap, one outlet of which is sealed 
imtil such time as it is used to admit the solution, and the other is 
connected via a trap bottle to a suction pump operated from the main 
water-supply. Alternative arrangements are shown in Fig. 5, p. 152. 


The first requirement for wood repairs is a good adhesive that is 
easy to apply and which will make a strong joint. Calcium caseinate 
is a powerful adhesive for wood; it was used in the Middle Ages for 
fiimiture-making and for joining planks together to make into broad 
panels as supports for painting. A good-quality hide glue is more 
convenient in practice, and can yield joints that are strong enough 
for most purposes. The tendency nowadays, however, is to choose 
one of the modem synthetic resin adhesives. These set by chemical 
action when mixed with a hardener to give very strong joints. Four 
main types^ are recognized, namely: 

Urea Formaldehyde (U.F.) adhesives. 

Phenol Formaldehyde (P.F.) adhesives. 

Resorcinol Formaldehyde (R.F.) adhesives. 

Epoxy resin adhesives. 

The first three tend to shrink on setting but the epoxy resins are 
characterized by the fact that they set without imdergoing any appre- 
ciable contraction in volume. 

The appHcation varies according to the manner in which the 
hardener is mixed with the resin before use. 

* These are marketed under a variety of trade names by different firms. DetaUs may be 
obtained from the following; British Industrial Plastics Ltd., i Argyll St., London, W. i; 
Aero Research Ltd., Hinxton, Duxford, Cambridge; Leicester Lovell & Co. Ltd., St. 
Christopher’s Works, North Baddesley, Southampton. 



(1) The hardener is mixed with the resin in the stated proportions 
before the adhesive is apphed to the joint. 

(2) The hardener is apphed to one side of the joint, the resin 
apphed to the other side, and the two held together till the 
adhesive sets, or 

(3) the adhesive is supphed as a dry powder consisting of resin plus 
hardener to which it is only necessary to add water before use. 

Detailed instructions are supphed by the manufacturers in each 
case. I In making a choice of adhesive it may be of significance that the 
R.F. hardener is neutral, whereas the others are acidic or alkaline. 

If, in addition to repairing a joint with an adhesive, it is necessary 
to fih irregular cavities in such a manner that the filler wifi itself 
confer strength, a modem gap-filling adhesive must be used, one 
of the most satisfactory being Aerohte 300, a U.F. resin. If it is 
merely a question of filling up a hole, choice may be made from a 
large selection of materials. The following are available — ordinary 
putty (whiting and hnseed oil); Alabastine (a fine hard-setting plaster 
of Paris) ; and a viscous celluloid^ syrup in which has been incorporated 
a porous filler such as sawdust or pumice powder; if a black filler is 
not out of place, Chatterton’s Compound may be used, apphed with 
a hot knife. Glue and sawdust may sometimes be all that is required, 
and awkward holes may be filled in with shaped balsa wood inserted 
with an adhesive. In pegged repairs of this nature, extra wood must 
always be allowed and the excess cut or filed away after the adhesive 
has finally set. 

Veneers and inlays tend to suffer from desiccation of the glue which 
causes them to hft; in this condition they are easily damaged with 
a duster and they should be temporarily secured with adhesive tape. 
This is a case where early attention is amply repaid. The process of 
laying back damaged veneer is a matter for the speciahst — amateur 
repairs are seldom very permanent or satisfactory— and for thejob to 

See also De Bruyne, N. A., and Houwink, R., Adhesion and Adhesives, 1951. 

Celluloid fillers must be applied in thin layers allowing time for the material to set 
between each application, otherwise shrinkage may be considerable. 



be done properly, it requires the use of a flexible glue^ and the appro- 
priate irons, the veneer being flattened and maintained under pressure 
until the adhesive has hardened. 

Marquetry can be kept from drying up by rubbing occasionally 
with a httle non-drying oil. Ohve oil answers well for the purpose. 
Almond oil is used for cleaning the wooden parts of violins and other 
musical instruments that are decorated with inlays. Wooden furni- 
ture having a good surface, whether French-pohshed or wax- 
pohshed, should be kept in condition by rubbing at intervals with a 
soft cloth. The original poHsh is always worth preserving whether 
apphed to veneer or sohd timber and it improves with time. Furni- 
ture polish coats the surface with a thin film of wax. This is a good 
thing, but wax should not be allowed to accumulate in cracks or 
crevasses as it will detract from the appearance and collect dirt. For 
cleaning pohshed furniture, of whatever kind, a suitable emulsion 
can be made by shaking together, vigorously, half a pint each of 
linseed oil, turpentine, and vinegar, to which is added a small tea- 
spoonfiil of methylated spirit. This mixture is comparatively 
inexpensive; it removes dirt and polishes in the same operation, and 
is harmless if appHed in moderation. 


Problems in preserving basketry are essentially those caused by 
mould growths and insect attack, and treatment is conducted on the 
same hnes as in the case of wood similarly affected. Basketry tends 
to become brittle, however, and very dusty. It may be cleaned by 
Lissapol froth, dried, and then impregnated with bleached beeswax 
dissolved in benzene. Celluloid in acetone forms a satisfactory 
adhesive for repair work. 

Bark-cloth garments disintegrate owing to the rotting of the 
stitching and they are then unable to sustain their own weight. 
Restoration is carried out by couching the bark to a Terylene hning, 
and an improvement in appearance results from spraying with a 

* Animal glue can be rendered flexible by the addition of starch derivatives, 
c.g. sorbitol. 


dilute methacrylate solution or with polyvinyl acetate spray solu- 
tion (p. 178). 


When wood is exposed for long periods in water or in wet clay 
or peat, it undergoes a form of decomposition arising from degrada- 
tion of the hgnocelluloses of which it is composed; much of the 
finer cellulose tissue disappears, but the thicker hgnin structures 
re main and these preserve the shape of the wood and often the general 
surface appearance. It is not uncommon to come across wooden 
objects from lake dwellings or from wells that appear to have sur- 
vived for generations with httle disfigurement, dark in colour but 
preserving their shape and ornament in minute detail. The loss of 
the finer cellulose tissue does not cause much alteration in the gross 
volume of the wood, but the porosity is much increased and the wood 
is able to absorb water like a sponge. When filled with water it is 
very heavy, and in this waterlogged condition it may be unable to 
bear its own weight so that if carelessly handled it may fracture across 
the grain. The preservation of waterlogged wood is one of the major 
problems for the archaeologist. Objects in this condition must be 
lifted on a rigid support or splint and packed at once in wet moss, 
cotton-wool, newspapers, textiles — anything that wiU keep them 
thoroughly wet — so that they can reach the laboratory in the con- 
dition in which they were found. They must then be transferred 
immediately to a bath of water, when they can be cleaned and 
examined at leisure, measured, and photographed. So long as the 
objects are kept wet they will keep their shape, but if allowed to 
dry without being specially treated, the weakened cellular tissue that 
is held apart only by water will collapse and the specimens wiU then 
be damaged beyond recovery. 

It goes without saying that waterlogged wood can be preserved 
in water (containing preferably a disinfectant such as 2 per cent, w/v 
of carbohc acid), but it is much more interesting if it can be dried 
without becoming deformed in the process. Two general methods 
are available for accomphshing this. The first is to replace the water 



in the pores of the wood with a hquid that will sohdify there and 
make contraction impossible as the wood dries, and the second 
method is to fix the tissue so rigidly that it will not coUapse after 
the water is removed. The alum process is an example of the first 
method, and the alcohol-ether-resin process an example of the second. 

I. The alum process 

The almn process depends on the fact that potash alum is a crystal- 
line substance freely soluble in hot water, but only soluble to about 
10 per cent, in cold. If a hot concentrated solution can be made to per- 
meate the tissue by soaking the wood in the hquid for sufficient time, 
the mass will sohdify as soon as the specimen is removed from the 
hot solution and no shrinkage will be possible. In practice, glycerine is 
added to the alum solution to facditate impregnation and it may help 
also to preserve the colour of the wood. 

The composition of the alum bath is important. Alum is dissolved 
in hot water in such proportions as to give the strongest concentra- 
tion consistent with reasonable fluidity at a temperature near the 
boiling-point of water. The solution is prepared in a copper or an 
iron vessel — not galvanized iron as the alum would dissolve the zinc. 
From two to three parts by weight of potash alum are dissolved in 
one part by weight of boding water to which has been added one 
part by weight of glycerine. The wood is submerged in this solution, 
using lead or iron weights if necessary to keep it below the surface, 
and the bath is maintained at 92-96° C. for at least ten hours, a longer 
time being allowed when the timber is bulky or has a close grain. 
Rosenberg^ recommends up to thirty hours — from which it will be 
seen that the treatment of heavy timber may take a long time. The 
level of the hquid should be maintained during heating operations by 
adding hot water as required. 

When it is considered that impregnation is complete, the wood 
should be hfted out, washed quickly with warm water to remove super- 
fluous alum from the surface, and set aside to cool. The residual white 
deposit which invariably remains is brushed off when dry, or removed 
‘ Rosenberg, G. A., Museums Journal, 193 3 > 33 > P- 432- 


with a cloth wrung out in hot water or preferably in hot glue size, 
which, with the alum, will form an insoluble skin. After a few days, 
when quite dry, the surface is sealed as described on p. 143. 

There is evidence that the best results are obtainable with hard- 
woods rather than softwoods and that success depends on the use of 
a sufficiently concentrated hot solution of alum, and on securing 
adequate penetration. An example of the satisfactory apphcation of 
the process to thin slats of yew is described in detail by Moss^ in 
the preservation of an Anglo-Saxon wooden bucket. 

2. The alcohol-ether-resin process 

The process of drying tissues in successive baths of ethyl alcohol to 
remove water and following this by immersion in successive baths 
of ether to remove alcohol is a well-known biological technique. If 
the final ether bath contains a resinous material in solution, the resin 
will remain in the wood on the evaporation of the ether, and the 
wood structure may thus he fixed rigidly. 

This has been found to be a very successful method of dealing with 
small objects. Werner Kramer has obtained excellent results by 
the alcohol-ether process, impregnating the wood finally under 
vacuum vsdth a concentrated solution of dammar resin in ether 
(Pi. 18). His success appears the more striking in that he bleaches the 
wood as a preliminary by immersing it in a 5 per cent, aqueous solu- 
tion of hydrogen peroxide for a week before commencing the alcohol 
treatment, and by this means the black shmy surface characteristic of 
waterlogged timber is converted to a pale slate or biscuit colour. The 
natural appearance of the wood may thus be restored and the grain 
revealed.^ Sometimes, as in the case of birch poles, it is even possible 
to recover traces of the silvery bark, the presence of which was un- 
suspected prior to treatment. Stone axe handles from a prehistoric 
site in Switzerland responded most successfully to the method and 
emerged as if they had been newly fashioned. 

* Moss, A. A., Museums Journal, 1952, 52, pp. 175-7. 

^ The method was developed at the laboratory of the Swiss National Museum, 
Zurich, under the guidance of Professor Emil Vogt. 



In recent work at Copenhagen, Christensen* has found that it is 
quite possible to dry waterlogged wood if the water is first replaced 
by alcohol, and this, in turn, by a liquid of low capillarity such as 
ether, which is then dried out very quickly by sudden exposure in 
a vacuum. This is done by having the object in a closed vessel con- 
nected by a stop-cock to a larger vessel that has been previously 
evacuated: the opening of the stop-cock reduces the pressure in the 
smaller vessel very quickly, and causes rapid volatifization of the 

Such vacuum-drying prevents shrinkage, but in thin woods may 
cause some warping. If a Httle shrinkage is permissible, the ether may 
be allowed to evaporate in the ordinary way without employing a 
vacuum technique and there is then no warping. 

The scheme advanced by Christensen is briefly as follows: 

1st Bath. Alcohol— in volume equal to five times the bulk of the 

2nd Bath. Repetition of No. i. 

3rd Bath. Fresh dry ether — Duration 2 days. 

4th Bath. Repetition of No. 3. 

Alcohol of at least 95 per cent, strength is used and great care is 
taken to ensure that the time of soaking will be adequate to remove 
aU water. For timbers such as oak and ash which are close-grained 
the time may be calculated at one day for every centimetre length 
of wood, but it is emphasized that the soaking in alcohol cannot be 

In normal cases the ether may be evaporated very quickly as de- 
scribed above, but when the object is thin and easily warped, the 
safer procedure is to allow the ether to evaporate naturally. Objects 
are then strengthened by impregnation in a 3 per cent, solution of 
polyvinyl acetate in pure benzene. The object is dipped in this 
solution for a moment only, and when dry, it is coated with a 10 per 
cent, solution of dammar resin in petroleum ether (B.P. 80-100° C.). 

‘ Christensen, B. B., Om Konservering af Mosefundne Trxgenstande, 1952 (National 
Museum, Copenhagen). 



This must be applied sparingly. It is unwise to attempt to cut short 
the process by going directly from the alcohol bath to impregnation 
with polyvinyl acetate, so dispensing with the ether, as this leads to 
unsatisfactory results. 

Both the Kramer and Christensen processes have been tried in 
the British Museum Laboratory and found satisfactory. The hydrogen 
peroxide method of pre-bleaching had been used for a niunber of 
years to reveal wood grain but always in a restrained fashion. More 
recendy, following a demonstration at Zurich, the bleaching was 
carried out in the laboratory to the limit and the specimen, a Roman 
scoop of hard wood, was restored to what must have been its original 
appearance. In this case the final bath was a strong solution of dammar 
resin in ether, apphed under vacuum. It was found that a piece of 
softwood treated in the same fashion could not be bleached beyond 
a pale slate colour, and it seemed to become more porous than the 
hardwood, but even in the softer wood the fineness of detail was 
well preserved: chisel marks survived and actually a series of fine 
saw-marks, no more than i millimetre apart. This bears striking 
testimony to the dehcacy of the method. 

While alcohol and ether are expensive and involve a serious fire 
risk, they are certainly to be preferred for the treatment of small 
objects, and there seems no doubt that this treatment coupled with 
hydrogen peroxide bleaching would be justified in all cases where it 
is essential to preserve fine detail of workmanship. 


An unusual type of restoration was successfully accomplished in 
the case of the iron sword discovered in its wooden scabbard by 
Sir Mortimer Wheeler in the excavations at Stanwick, Yorkshire, 
in 1952. This was a umque object dating to the first century a.d. 
(Pi. 19). As it was waterlogged, it was dispatched by the first train 
to London packed in a mass of wet newspapers, and it was this 
prompt action that made its preservation possible. Reference is made 
later (p. 282) to the corrosion of the iron sword by sulphate-reducing 



bacteria and to the preservation of the -weapon. The following 
details relate to the preservation of the wooden scabbard. 

The scabbard was made from two slats of ash of a maximum thick- 
ness of I inch and tapering towards the edges. These were held to- 
gether by a series of bronze rings of symmetrical eye-shape set at 
intervals, and decreasing in size from the mouth-piece to the chape. 
As the sword could not be withdrawn from the scabbard, in which 
it was firmly held by its iron sulphide corrosion products, the alum 
method could not be used for preserving the wood. This acidic sub- 
stance in reacting with the sulphide might have caused disintegration 
beyond control. It was decided, therefore, to remove the scabbard 
from the sword by dissection under water. The varying width of the 
bronze fittings afforded the possibiUty of their removal from the 
chape end; this proved to be a dehcate but not a difficult task, and 
the hronze mouth-piece was finally released from its own end. 

After the removal of the hronze fittings the wood was detached 
with the aid of scalpels from one side of the sword-blade and floated 
on to a glass plate cut to the length of the scabbard. The blade was 
then reversed and the remaining wood removed and floated on to a 
second glass plate. 

The specimens were then treated by immersion in a series of 
alcohols of increasing strength in order to eHminate the water. The 
treatment was carried out in boxes made specially from tin-plate. 
The alcohol in the wood was eventually replaced by amyl acetate, 
the final immersion being caried out in a solution of celluloid dis- 
solved in equal volumes of amyl acetate and acetone to a strength of 
about 5 per cent. Throughout these operations, the wood remained 
on the glass-plate supports and was not handled until it had been 
strengthened. This work took about fourteen days, and on the 
evaporation of the solvent, the thin porous wood was found to be 
toughened quite satisfactorily. 

For the reassembly of the scabbard it was necessary to prepare 
a core of such dimensions that, with the old wood in position, the 
bronze rings would shp on from the pointed end, and the recon- 
structed scabbard would conform to original dimensions. But here a 



difficulty arose due to minor irregularities in the shape of the metal 
fittings. This necessitated the use of a resdient core which was made 
as follows : a piece of wood was cut sHghtly thinner and narrower than 
the sword, and to this was fiixed along its length a piece of thin steel 
cut from a flexible steel rule, the attachment being made with a ban- 
dage soaked in shellac. This core was then impregnated with a solu- 
tion of nitrocellulose and the old wood of the scabbard was attached 
to it on both sides with Durofix, the contact being made as intimately 
as possible. The rings could now be shpped on from the pointed end, 
there being some sHght elasticity in the core in consequence of the 
action of the spring in its interior. When the adhesive hardened the 
core became rigid and some minor repairs with a mixture of celluloid 
and sawdust completed the restoration. 

It was, of course, never intended to return the sword to its scab- 
bard; the two were subsequently mounted side by side on a wooden 
baseboard to form an exhibit, interesting in itself and convenient for 
purposes of study. 


The preservation of heavy waterlogged timbers, such as dug-out 
canoes, presents a special problem because of the difficulty of trans- 
port and the expense of the large quantity of materials required for 

The wood is cleared of mud and drawn out of the water or moist 
ground on an improvised stretcher or sledge preferably of sheet metal, 
and it is set in such a position that it can be washed down, measured, 
and photographed. The usual procedure is then to cover it under a 
good heap of earth and to leave it for at least a year but preferably 
longer so that the water will drain gradually from the timber. When 
this takes place slowly the heaviest oaken timbers do not warp, 
although the surface may, to some extent, tend to open up. In the 
case of a dug-out canoe, the sides, which are much thinner than the 
base, will tend to warp inwards and these must be supported before 
covering with earth by inserting wooden cross pieces padded Avith 
sacking, and lashing them in position. These should be readjusted 



when the earth is finally removed and before the vessel starts to 
undergo acclimatization to atmospheric conditions. The acclimatiza- 
tion process may take an additional year or longer. During this time, 
the wood must be protected fiom the weather, particularly from 
frost and direct sunshine. Towards the end of the acclimatization 
period, the surface may be sealed by painting it with a solution of 
sodium sdicate of high sUica content diluted with distilled water, e.g. 
P 84 brand of sodium sdicate^ diluted with six times its volume of 
distilled water. Such treatment has proved to be satisfactory in several 
cases. Sometimes where the wood has appeared to dry well impreg- 
nation is unnecessary, but with large thin timbers the element of 
chance is considerable and much depends on the grain of the wood 
and its freedom from knots. AU would agree, however, that the 
slower the process of drying the better the chance of achieving a 
successful restoration. It is, of course, fatal to bring such an object 
indoors, at least until it is in equihbrium with conditions out of doors. 
The proximity of radiators, for example, can imdo the work of years 
expended in caring for the object in the early stages. 

The remains of two Iron Age boats were recovered from the 
River Humber at North Ferriby in 1947. These consisted of keel 
planks and major portions of the lower timbers of the vessels sewn 
together with thin yew branches and caulked with moss. The boats 
were salvaged by lashing them to metal skids at low tide and drawing 
them up to the river bank out of the mud. As these early rehcs were of 
unique interest, they were taken at once to the National Maritime 
Museum, Greenwich, where they underwent a course of treatment 
to dry the wood with the minimum of distortion. 

Previous investigations by the finders^ had yielded several old 
spars, some of which had been placed in glycerine and some allowed 
to dry naturally, thus providing useful material for comparative 
study. While the air-dried wood showed httle deformation, there 
was a definite tendency for surface layers to develop a craquelure 

‘ Imperial Chemical Industries Ltd. 

^ Wright, E. V., and Wright, C. W., ‘The North Ferriby Boats’, Mariner’s Mirror, 
1947. 33 . No. 4, October, p. 235. 



which on the thinner portions formed a system of raised flakes. A 
long-term experiment with a waterlogged spar showed that glycerine 
could be introduced to replace the water in sufiicient amount to slow 
down the rate of drying, leaving a surface that could eventually 
be sealed with shellac; by this means the bulk of the water was 
ehminated without the formation of craquelure. 

Galvanized iron tanks of ample size to accommodate the timbers 
were constructed by bolting smaller units together and these were 
provided with drain holes and plugs so that the final de-mudding 
could be carried out in the tanks. The wood was arranged in a single 
layer supported above the bottom on bricks. After washing and 
draining thoroughly, the plugs were inserted and the wood was 
covered with a dilute glycerine solution which was replaced three 
or four times over a period of two years, the progress of operations 
being followed by stirring up the solution at intervals and testing 
the density of the hquid with a hydrometer. When the glycerine 
hquors corresponded in strength to those from which the test spar 
had been taken, they were run off and the timber was allowed to dry 
in air for more than a year. The pieces were then removed and treated 
individually with a solution of shellac in alcohol. With heavy timbers 
there must always remain an element of doubt as to when they are 
dry enough to be considered stable, and while in the present instance 
there appear to have been no comphcations that were not immedi- 
ately overcome, it is recognized that it may take years for equflibrium 
to be estabhshed and it will be necessary to keep the work under 
observation and to apply more shellac if the wood requires it. In 
the present instance even the caulking and yew-stitching survived 
treatment and all the timbers were decidedly improved in ap- 

The largest water-logged vessels that have been preserved are un- 
doubtedly the Viking ships at Oslo.^ There are two superb examples, 

■ Brogger, A. W., Falk, H. J., and Schetilig, H., Oseberg Fundet, Vols. 1-5; Vol. i 
(1917) gives technical details of the restoration work. See also The Viking Ship Finds, 
Oslo Museum (i937)> 2itd Dannevig-Hauge, T., Konservering av Tre, Universitetets 
Oldsaksamlings Arhok, 1949-50, Oslo, 1951. 



the Oseberg ship and the Gogstad ship. The third, the Tune ship, 
which is in a fragmentary condition is still of much interest in show- 
ing clearly how such vessels were constructed. These ships, and many 
of the carved wooden objects that they contained, were preserved 
mainly by the alum method. Before beginning treatment exact sizes 
and shapes were recorded by taking plaster impressions while the 
objects were stiU wet. The alum-treated wood was finally coated 
with a mixture consisting of equal volumes of turpentine and linseed 
oil with the twofold object of improving the appearance and water- 
proofing the surface. 



From earliest times bone and ivory have exerted a peculiar fascina- 
tion for the craftsman. Bone was the more easily come by and, as it 
was available in a variety of shapes and degrees of hardness, selected 
pieces could be fashioned into all sorts of useful objects — fish-hooks, 
arrow-heads, tools, and implements. Ivory, though of standard shape, 
was the more suitable material for fine work and lent itself to decora- 
tive treatment. It could be carved, etched, stained, painted, gilded, 
inlaid with metals and with precious and semi-precious stones; it 
could be used, moreover, for inlaying in wood, and for veneering. 
When a material was required for the finest carving, the choice fell 
on ivory rather than on bone. Thus, ivory was in demand for the 
plaques and reUefs that were used for diptychs and bookbindings. It 
was also ideal for carving figurines and statuettes, and specimens are 
extant that date back to the earhest dynasties of Egypt. A great 
wealth of ivory-carving has survived from early Christian times in 
the West, and many fine works in ivory have also come down to us 
from remote civilisations in China and the Near East. Although bone 
was more generally used for utihtarian purposes, it was often brought 
into service for ornamental purposes as well, owing to the relative 
scarcity of ivory. 

When worked and decorated, the two materials may be so much 
alike that they cannot be distinguished by mere inspection, and 
identification is then a task for the speciahst. It is not possible by 
chemical means alone to distinguish bone from ivory. In both 
materials the main inorganic constituents are the same, namely 
calcium phosphate associated with carbonate and fluoride, and the 
organic tissue of both is ossein; while the latter varies in quantity, it 
is usually present to at least some 30 per cent, of the total weight. 

SWOHD AND WOODl.N S C A li II A l( D I ROM S I' A N \V I ( ' K 


Before treatinent 

Sec Jivniispiccc Jor coihiitioiis pfrer tiwirincin 



Microscopic examination will, however, permit of a differentiation, 
since bone and ivory both have a cellular structure, and the micro- 
structure of each is definitive. In cross-section, bone shows a rather 
coarse grain with characteristic lacunae, whereas ivory, being com- 
posed of the hard dense tissue known as dentine, is more compact 
and is characterized by the presence of a network composed of tiny 
lenticular areas resulting from the intersection of systems of striations 
that may be seen radiating from the centre of the tusk. In favourable 
circumstances this can be observed without sectioning by studying 
the surface with a pocket lens. 

The main factors bearing on the conservation of bone and ivory 
may be deduced from a consideration of their physical structure and 
chemical composition. Bone and ivory are anisotropic having 
directional properties and for this reason they are easily warped upon 
exposure to heat and damp; furthermore, they are decomposed by 
the prolonged action of water due to hydrolysis of the ossein, and 
the inorganic framework is readily disintegrated by acids. Being 
porous and of hght colour, bone and ivory are easily stained. They 
tend to become brittle with age and they lose their natural colouring 
when exposed to surdight. When burnt, they become grey or blue- 
black,^ and when buried in the ground for prolonged periods of 
time they are greatly weakened and may be disintegrated either by 
salt incrustations or by water; in the latter case they are converted 
ultimately to a sponge-hke waterlogged material similar to water- 
logged wood. Under other circumstances they may become fossdized. 
With the onset of fossihzation the organic content gradually dis- 
appears and the remaining calcareous matter becomes associated with 
sihca in the form of quartz and with mineral salts derived from the 
ground. When bone and ivory are in good condition it is possible 
to preserve them by taking very simple precautions, but when they 
are in a fragmentary state, waterlogged, or fossihzed they can only 
be cleaned, strengthened, and stabUized; a satisfactory restoration 
may be impossible. 

Two general features should be considered before discussing 
* Bone-black and ivory-black are two well-known pigments. 

B 8157 




detailed methods of preservation and restoration. These relate to size 
and colour. Bone and ivor)’^ are usually only available in pieces of 
hmited size, so that large objects made from these materials will of 
necessity be composite in structure. In the case of ivory, for example, 
when a large object does not conform to the shape of the tusk, the 
artist may add portions by doweUing or jointing (tongue and groove, 
half-lap, &c. ) and, when any warping of the component pieces occurs, 
the greatest strain will tend to be imposed on the joints. This is parti- 
cularly noticeable in the case of books that have covers made of 
ivory; a clasp or even a pair of clasps may have been fixed to keep 
the covers shut and the pages flat, but should swelling of the pages 
occur (and this is by no means unusual if they are of parchment, see 
p. 47), the strain on the ivory covers may be considerable and the 
bindings may open along the joints. This type of accident is caused 
by the cumulative effect of damp which causes the pages to swell, 
but surular accidents may be the direct result of exposing the ivory 
to sunhght or to heat. Joints may open, veneers may warp, and if 
warping is serious, it may be impossible to restore the original shape. 

Old ivory often has a yellow colour and this is accepted as a form 
of natural patination which may help to enhance the appearance. On 
the other hand, coloration may have been artificially apphed, as is 
often done by the Japanese carver of netsuke, who displays the 
greatest originahty in arriving at a surface effect that will show his 
handiwork to the best advantage. The greyish colour of burnt ivory 
has already been mentioned, and in some of the Nimrud ivories there 
seems reason to suppose that the burning was intentional and was 
carried out with the object of achieving a particular colour effect. 
Staining may be due to a residue of some former embeUishment 
executed in a fugitive dye-stuff that has decomposed; in this case the 
residual colour may show some signs of regularity or of pattern work 
which may be of interest and value to the speciahst as evidence of a 
decoration that once existed but no longer survives. There also arises 
the problem of adventitious staining which may be due to a variety 
of causes. When there is no sign of regularity in the staining its origin 
may be obscure; it may possibly be due to some photochemical 


effect or to contact of the bone or ivory with another material. Rust 
stains and copper stains can be helpful in the reconstruction of a 
broken object by showing which pieces originally fitted together, or 
in caUing attention to the previous existence of hinges or metal 
attachments. In the case of the Franks whalebone casket in the British 
Museum, staining of this kind was particularly valuable to those 
engaged in its reconstruction. Finally, there may be disfiguring stains 
arising through use or accident. Stains may he removed by appro- 
priate chemical treatment, but since there are so many possible causes 
of coloration in bone and ivory objects, the greatest caution is neces- 
sary before undertaking any treatment that might result in the loss 
of important characteristics; only in exceptional cases can the removal 
of stains be justified, because although chemical treatment may dis- 
charge the stain, it may result in damage to the texture and patina of 
the ivory, and thus magnify the disfigurement. 


Removal of surface dirt 

Having referred to the damage that can be caused to bone and 
ivory by exposure to water, it may seem illogical to assert that wash- 
ing with soap and water can be carried out without danger. This can, 
however, be done if the objects are in good condition, and if they 
are dried with soft towelling immediately after treatment, so that the 
wetting is confined to the surface. The use of water containing a 
detergent (soap, or one of the new surface-active agents) is indeed 
essential for the removal of soot and grease, but the object must not 
be allowed to remain in contact with the aqueous solution longer 
than is absolutely necessary, and for this reason the washing is carried 
out with a brush rather than with cotton- wool or a sponge. The hard- 
ness of the brush and the length of the bristles is also important; too 
soft a brush would only prolong the time necessary for cleaning. 
The brush must therefore be selected with some care to suit the job 
in hand, taking into account the strength of the object and the 
amount of grime to be removed. 



A more complicated problem arises in the case of partially decayed 
ivory or bone, especially when the poHshed surface skin has deterior- 
ated and there is evidence of cracking. In this case the ivory is washed 
as before, but instead of being dried with towelling it must be passed 
immediately through two or three changes of 95 per cent, alcohol 
before being dried by contact with absorbent material such as a paper 
handkerchief or blotting-paper. 

Removal of salts 

One of the most difficult problems has been that of cleaning salt- 
incrusted and semi-fossihzed ivories from Egypt. These may be very 
absorbent and, if placed in water even for a few minutes, they tend 
to spht and to warp, providing an almost impossible task for the 
repairer. However, the cleaning of such ivories ceased to present any 
real difficulty when it was discovered that it was not necessary to 
remove all the salt but only the surface deposits, and that success de- 
pended on the speed of operations. In a typical instance a cracked 
ivory, disfigured by a salty incrustation, was placed in a shallow 
earthenware dish, and covered with distilled water. This water was 
discarded and replaced with fresh distilled water at intervals of five 
seconds, four or five washings being considered to be sufficient. The 
ivory was then immediately washed in 80 per cent, alcohol for thirty 
seconds, followed by two washings in 95 per cent, alcohol or 
industrial methylated spirit for a similar period. Finally, the object 
was immersed in ether for one minute, and dried in air. By this 
procedure the entire cleaning was completed within three minutes, 
the ivory having been exposed to water for a much shorter time, and 
it survived without any apparent change of shape or extension in the 
cracking. Such treatment may seem to be too quick and easy to have 
any permanent effect, but, in fact, it has been apphed on many occa- 
sions in deahng with salty ivories from Egypt and Palestine. The 
results have been entirely successful, and the times have been cut 
down well below those stated where specimens were particularly 
frail or fragmentary. 

Incrustations of carbonate provide yet another complication. In 



this case, the use of acid is essential to decompose the carbonate, but 
during decomposition the effervescence due to the evolution of 
carbon dioxide gas would inevitably result in disintegration of the 
ivory. This difficulty is overcome by applying the acid locally only, 
i.e. no more than a square centimetre is dealt with at a time. Ivories 
incrusted with carbonate are cleaned rmder a binocular microscope 
at about 10 diameters magnification, applying a i per cent, solution 
of hydrochloric acid with a small water-colour brush repeatedly. As 
effervescence ceases, the excess of liquid is mopped up either with a 
second brush or with shreds of blotting-paper apphed with forceps. 
The acid must be used in quantities just sufficient to soften the incrus- 
tation, which is then removed with a needle. Finally, in order to 
eliminate the last traces of acid, the ivory is washed in several changes 
of distilled water for a few seconds at a time, and then dried by 
the alcohol and ether method already recommended for dehcate 

Occasionally one comes across an ivory^ that is incrusted with 
calcium sulphate (selenite). This incrustation cannot be removed by 
chemical means ovting to its comparative insolubiHty; any attempt 
to do so would involve a long soaking process which would cause 
the ivory to disintegrate. The only procedure that can be recom- 
mended for removing calcium sulphate from ivory is to reduce the 
incrustration mechanically by the use of dental equipment — inverted 
cones and burrs. Needless to say, such treatment can only be under- 
taken if the ivory object is sufficiently robust to withstand the 
mechanical stresses involved. 

In describing the different methods of cleaning bone and ivory, 
emphasis has been laid on the importance of cutting down the time 
during which they are exposed to the action of chemicals and even 
of water, but in an exceptional case it may be necessary to break 
these rules in order to save a fine specimen. Such action is only justi- 
fiable when the risks to be taken are fuUy understood, so that treat- 
ment can be suitably modified to reduce these to a minimum, and 
when there is a reasonable hope of success. An interesting example is 
provided by the restoration of a well-known Egyptian ivory figurine 



in the British Museum, once thought to represent Menes of the First 
Dynasty.^ This ivory had been broken in antiquity; the parts were 
incrusted and held together with calcium carbonate which, in 
crystaUriing, had opened up the structure to form a series of strata 
interspersed with carbonate, so that the object was both swollen and 
deformed. In order to restore the object to its original shape it was 
necessary to clear away all the carbonate from the cracks; therefore 
the cleaning could not be confined to the surface as is normally the 
case. Actual soaking in dilute hydrochloric acid followed by washing 
was necessary. It was foreseen that effervescence during the acid 
treatment would probably cause the specimen to disintegrate, and 
this eventuahty was prevented by previously impregnating the ivory 
under vacuum with a solution of celluloid (2 per cent.) and allowing 
it to dry. The celluloid solution had the effect of slowing down the 
penetration of the acid, thus holding the effervescence in check. When 
the calcium carbonate had dissolved, the constituent pieces fell apart. 
These were washed in distilled water, dried through alcohols to ben- 
zene, and then impregnated with a solution of dammar resin dissolved 
in benzene, using the vacuum technique already described. Re- 
assembly of the fragments, using a viscous solution of dammar resin 
in benzene as the adhesive, presented no problem and the original 
shape was recovered. After restoration the figurine was found to be 
more complete in regard to its ornamentation than had been pre- 
viously suspected. Only one major gap remained and this was filled 
in with a mixture of beeswax and camauba wax and tinted with thin 
oil-paint to match the biscuit colour of the ivory. It should be noted 
that when cracks in ivory arise from warping it is not desirable, in 
general, to fill them in, because under different environmental con- 
ditions they may tend to close up, and if this natural tendency for the 
cracks to close is prevented by the presence of a filler, it may cause 
the ivory to rupture. However, in the present instance a small portion 
of ivory was missing, leaving a fissure that caused a structural weak- 
ness. This weakness was overcome by the filling which, at the same 
time, had an aesthetic value in nunimizing the disfigurement. 

’ Figure illustrated in Plenderleith, H. J., The Preservation of Antiquities, 1934. 




When it is merely a question of consolidating a dry powdery ivory, 
several suitable transparent media are available, notably polyvinyl 
acetate and the polymethacrylates. Polyvinyl acetate lacquer was 
recommended by the present author as long ago as 1934 as a medium 
for impregnating and strengthening bone and ivory and also in 
stronger solution as a cement.^ Todd^ has described how he consoh- 
dated carved ivory spatulas of the Shang dynasty with Gelva (a poly- 
vinyl acetate) by impregnation under vacuum. Similar materials 
have been used for strengthening bone.^ The use of an aqueous 
emulsion of polywinyl acetate for hardening bone that has been 
excavated from bogs has been described by Purvis and Martin."^ 
Waterlogged bone provides a much easier problem than water- 
logged wood, as it can be strengthened directly by vacuum impreg- 
nation with dilute polyvinyl acetate emulsion in an apparatus such as 
that illustrated in Fig. 5. Purvis and Martin state that the individual 
globules of such an emulsion have a diameter of about o-oi /i. When 
the pressure is released, the smallest interstices of the bone are filled 
with the emulsion and it only remains to drain the bones on a wire 
frame. The treated material is described as being robust and as having 
recovered its original colour variations and minute surface archi- 
tecture, and as being capable of satisfying the requirements of the 
most exacting osteologist. McBumey® speaks very favourably of the 
excellent quahties of polyvinyl acetate (presumably the emulsion) 
for hardening bones during excavation both before removal from 
the ground and immediately after, especially when the bones are 
very wet and soft. Bedacryl (a polymethacrylate emulsion) also 
works well, and this may be readily transported as a concentrate for 
field work and diluted with water to the required consistency before 

‘ Plenderleith, H. J., op. cit., p. 19. 

^ Todd, W., Technical Studies in the Field of Fine Arts, 1940, 9, p. 160. 

^ See ‘Note on the use of Polymerised Vinyl acetate and related compounds in the 
Preservation and Hardening of Bone’, Am.J. of Phys. Anth. 1936, 21, pp. 449-50- 

'' Purvis, P. E., and Martin, R. S. J., Museums Journal, 1950, 49, p. 293. 

® McBumey, C. B. M., Arch. News Letter, 1954, 5, p. 57. 

* Imperial Chemical Industries Ltd. 


use. These emulsions are stable so long as they are protected from 

Although nitrocellulose lacquers and adhesives suffer from certain 
minor defects, these are outweighed by the advantages that they 

Fig. 5. Apparatus for Vacuum Impregnation 
A. Impregnating Medium containing the Object. b. Vacuum Vessel, 

c. Trap with Pressure Gauge. D. Water pump attached to mains supply. 

offer for strengthening and repairing bone and ivory — ease of 
apphcation, rehable adhesive properties, and ease of removal. The 
latter characteristic is important as it allows adjustments to be made 
in the course of repair work, and makes it possible at any time in 
the future to remove the adhesive, should this be desired. By using 
Durofix, a comphcated repair like that of the St. Cuthbert comb, 
which had spht into countless fragments, could be tackled with con- 
fidence and with the knowledge that under museum conditions the 
restoration would be, to all intents and purposes, permanent. 


Particular care must be taken in handling ivories when there are 
features of interest that must be preserved ; for instance the carved 
ivory plaques of the Christian era that have prayers written on the 



back in ink, and others that are coated with wax as a ground for 
writing with a stylus. There are cases where damage has been caused 
in the moulder’s shop when making plaster repHcas of ivories. 
Moulding is not necessarily dangerous provided the moulding 
materials can he used cold. The temptation to use a warm jelly mould 
or warm ruhher latex, which is so convenient for dealing with under- 
cuts, must be resisted, as such treatment may cause cracking or stain- 
ing. A safer procedure is to piece-mould the ivory or, preferably, to 
employ a cold jelly mould prepared from alginate;' but all materials 
that require to be used wet necessitate pre-treatment of the ivory to 
protect it from possible damage by water. This is generally done 
by giving the object several thin coats of nitrocellulose lacquer 
diluted with acetone. Such a thin coating need not affect the sharp- 
ness of the rephcas, and may easily be removed afterwards by swab- 
bing with solvent. Perfect rephcas may be obtained by the use of 
alginate and, when expertly handled, this material imposes no strain 
on the ivory. 


During an excavation at Nimrud in 1952, Professor Mallowan 
recovered a number of magnificent ivories from the bottom of a 
well, dating from the period of Assur-nasir-pal II (883-859 b.c.) and, 
by drying them slowly and uniformly, he was able to preserve their 
shape and prevent serious cracking. It may be of interest to describe 
the subsequent treatment that was apphed to one of them — a master- 
piece of exquisite carving, embeUished with gold and incrusted with 
lapis lazuh and carnehan (Frontispiece). 

When this reached the British Museum Laboratory it was covered 
with fine clay (Pi. 20), and an X-ray examination showed the pre- 
sence of deep cracks which were widest at the back, or external side, 
of the tusk, but, fortunately, scarcely apparent on the decorated side. 
The clay was carefully scraped from the back and, as this revealed 
a surface without decoration, it was rigidly secured (as a first-aid 
precaution to prevent any further opening of the cracks) by backing 

' Alginate Industries Ltd., Dipple by Girvan, Ayrshire. 



it with layers of broad adhesive tape, the equivalent of surgical 
strapping.^ The front was then dealt with, working under a binocular 
at a magnification of lo diameters. First the upper layers of clay were 
removed with needles to expose the sculpture. Many fragments of 
gold leaf were recovered from the clay; these were washed with 
I per cent, nitric acid and then with water, and set aside for subse- 
quent replacement. The ivory was further cleaned with pellets of 
blotting-paper held between pointed forceps and moistened with 
detergents in order to peptize and release the clay with the minimum 
of strain, and later the surface was washed and pohshed by the same 
technique, doing small areas at a time. The final stage in the restoration 
was the replacement of the loose fragments of gold leaf in their 
correct positions on the cleaned ivory, using Durofix as the adhesive. 

The background of this superb object — thought to be part of a 
throne — is decorated with an all-over floral pattern consisting of 
alternating flowers and seed capsules, the stems and sepals being 
covered with fine gold. The forms are deeply carved and the thin 
walls that outline the flowers are also gilt so that they appear hke 
metal cloisons framing the lapis lazuh inlays of the petals. The seeds 
(or buds?) are represented by pohshed camehans of dome shape, 
serrated at the base to engage with the gilt ivory calyces. This rich 
background forms a canopy for the main carving below, a scene in 
high rehef depicting a honess in the act of killin g a Nubian. The ivory 
body of the animal is unadorned, and powerfully modelled. It 
stands out in sharp contrast to the relaxed human victim with his 
gleaming golden loin-cloth, and spikelets of crisp curly hair, an 
effect obtained by fixing gilt-topped pegs into the head which was 
stained black beforehand. In spite of the loss of much of the gold 
overlay and the blue and red incrustation, this ivory carving still 
gives the effect of a faceted polychrome jewel. 

The following techmcal points relating to the construction of the 
ivory were noted during the cleaning and are of special interest. 
The lapis lazuh inlays are of varying thickness, but the recesses in 

' Lassoband, Smith and Nephew Ltd. (Technical Tapes Division), Welwyn Garden 
City, Herts. 



the ivory are of uniform depth. Consequently, it was found neces- 
sary to bed the inlays in a cement so that their pohshed surfaces would 
be level with the cloisons. This is the only case known to the author 
where a blue cement has been used for inlaying lapis lazuli. It is com- 
posed of lime putty coloured to match the lapis lazuh by the addition 
of copper frit and where inlays are missing the ivoty^ is often stained 
by a residue of this frit cement. Another interesting observation was 
made on examining the back of fragments of gold leaf from the ivory. 
Traces of a brownish film were clearly visible on the gold under the 
microscope. This material which was of organic origin and swelled 
in water, becoming very sticky, appeared to be the original adhesive 
employed in laying the gold leaf. It is difficult to beheve that a rever- 
sible coUoid could have survived in such circumstances but appar- 
ently this was the case, and it was due no doubt to the protection 
from air and moisture afforded by the coherent film of gold. 


Many of the methods that have been described for the treatment 
of bone and ivory apply with sHght modification to kindred organic 
materials that are frail or porous, such as tortoise-shell, horn, and 
antler. These may be washed in the same way as ivory. 

In the case of horn, which is usually thin and hollow, the surface 
may be strengthened and prevented from flaking by impregnation 
with polyvinyl acetate and it may be possible to confer added 
strength by attaching a plastic tape or bandage to the inside surface 
of the horn. Only part of the interior can be covered, if a horn is 
intact, but this will be the thinnest part and the part, therefore, that 
has the greatest need of reinforcement. 

Antler usually has a spongy central tissue covered with a con- 
tinuous outer rind. Under conditions of prolonged burial the spongy 
tissue generally decays and only the outer rind survives. Thus the 
antlers of the Sika deer that once crowned a Chinese wooden deity 
were found to have lost all their spongy tissue; they were hollow and 
partially collapsed. The surface skin had survived, however, and it 



was possible to support this with a wire core bound with cotton- 
wool and reinforced where necessary with a filling of nitrocellulose 
and cork dust. The antler picks or levers found in flint workings in 
the chalk are usually dry, and in this case may be treated by pouring 
dilute polyvinyl acetate lacquer through the porous tissue. It should 
be noted that this may leave an unpleasant shine on the outside of the 
antler unless the surface is swabbed with toluene and the antler 
allowed to dry in toluene vapour. This can easily be done by sus- 
pending it in a tall vessel containing a couple of ounces of the hquid 

Although jet and amber are not chemically related to bone and 
ivory, their treatment may be referred to here as they are washed 
according to the methods outlined above for ivory. While fossil 
amber is insoluble in organic solvents it is not easily distinguished in 
appearance from other resins that are soluble and for this reason it is 
advisable to avoid the use of alcohol and ether altogether in cleaning 
objects of resinous origin. 

Finally, a special warning is necessary in dealing with jewellery 
that may be inlaid with materials such as shell, mother of pearl, and 
coral (which consist essentially of chalk). These substances are 
decomposed by acid; where there is any doubt about identifica- 
tion, cleaning-agents of an acidic nature must at aU costs be avoided. 



Easel paintings are defined as paintings on panel or canvas in 
tempera or oils, and they present unique problems in conservation 
because of their variety and complexity of structure. The pictorial 
image is, however, the important thing — colour and optical quahty 
have to be studied in relation to conservation — and it is for this 
reason that the care of valuable paintings is regarded as being the 
most exacting and responsible of all museum curatorial activities. 
Restoration is a matter for the expert as it involves an appreciation 
of aesthetic quahty as well as a specialized training lying outside the 
field of ordinary museum work; but the curator must himself be 
fanuhar with all aspects of his subject, and the present chapter is 
designed to provide such technical information as may help him to 
appreciate the fundamental factors involved in the care of paintings 
whether in the gallery or in the restoration studio. It deals with the 
structure of paintings, with the ills to which they are subject, and 
with methods available for their conservation. 


An easel painting cannot be regarded merely as a coloured surface. 
Although painting techniques may have changed through the cen- 
turies, the basic fact remains that, from the physical point of view, 
a painting is a stratified structure built up in a series of layers each 
of which may exhibit complexities of its own as shown diagram- 
matically in Pi. 21. There is first of all the support, a prepared panel 
or canvas; on this is spread the ground, which consists essentially of 
a white inert substance in a glue medium. On Itahan panels the inert 
substance was usually calcined gypsum (Ital. gesso), but Northern 



painters favoured the use of chalk. For the appUcation of gold leaf, 
a special ground was used in which the inert constituent was a species 
of clay known as bole. The paint layer overhes the ground and it 
consists of an aggregate of pigment particles in a binding medium. 
In tempera painting ^ the medium was egg yolk or the whole egg, 
but later, drying oils were more commonly used. Finally, there is 
usually a layer of varnish and this surface coating serves a twofold 
purpose: it protects the paint layer against atmospheric contamina- 
tion and also confers upon it an enhanced briUance. When a picture 
is in a sound condition, its various layers are fused and interlocked, 
the closest study reveaHng no sign of cleavage 

The character of the painting is mainly influenced by the nature 
of the binding medium and, as the medium carries the whites and 
blacks and other pigments, its stabihty is crucial to the hfe of the 
picture. By the test of time, egg-tempera and oils have proved them- 
selves satistactory media for easel paintings, but they suffer in varying 
degrees from a tendency to shrink and to become brittle with age. 
When a picture reaches a certain age, its surface usually becomes 
patterned with a series of micro-cracks which have picked up foreign 
matter and reveal themselves as dark hair-like hnes most readily 
discernible in the hghter passages; these hnes intersect and form a 
comphcated network known as craquelure. Many factors contribute 
to the general appearance of craquelure, but in the main it is due to 
the natural shrinkage of the medium on ageing and the inabflity of 
the brittle aged film to withstand the shght rhythmical movements 
of the support.^ The forces thus imposed on the paint layer may 
cause a craquelure which extends through both the painted layer and 
the ground; different types may be recognized as characteristic of 
painting on wood panel or on canvas. 

While superficially the craquelure on a painting may often appear 

Cennino Cennim has described the whole process of tempera painting in his famous 
Treatise publRhed in 1437 (date of earhest surviving text). Cennino D’ Andrea Cennini: 
II Lihro Ml’ Arte {The Craftsman’s Handbook), translated by D. V. Thompson, jr., 


^ Movements in response to changes of the relative humidity of the atmosphere 
(see later). 



to resemble the intentional crackle of a glaze on oriental pottery or 
porcelain, it differs in this respect that it is an acquired characteristic 
and develops slowly; it is not regarded as a disfigurement, and rarely 
is it of such a nature as to introduce instabihty. 

Two subsidiary types of craquelure may be mentioned. One is 
due to faulty technique, such as the appHcation of a layer of quick- 
drying paint over a slow-dr)ing paint; this type of so-called ‘youth’ 
craquelure is usually restricted to certain passages of the painting. 
The other condition is due to a faulty choice of material and is known 
as ‘alhgatoring’, a term used to describe the system of wide depres- 
sions in the paint layer resulting from the use of bitumen or asphal- 
tum, a material that is soluble to some extent in the binding medium 
and has notoriously bad drymg properties. 

In contrast with craquelure which is a more or less normal con- 
dition in old pictures, cracking and cleavage are serious forms of 
deterioration which affect canvases as well as panel paintings. A crack 
has the appearance of a crevasse in the paint layer, running inwards 
at right angles to the surface, and it usually penetrates through all 
the layers of the picture. It is caused by flaws in the ground or sup- 
port. Cleavage, the more insidious form of deterioration, is due to 
loss of adhesion between the superimposed strata of the picture; it 
manifests itself in the formation of blisters which, if they be 
allowed to fracture, will lead eventually to the loss of paint. 

While cracking and cleavage may be found in pictures of any age, 
they are to be seen most frequently in old pictures where the bind- 
ing media have lost much of their original elasticity as a result of 
chemical changes, and are no longer able to adapt themselves to 
movements of the support occasioned by exposure to environ- 
mental changes. Both these conditions are revealed particularly 
when there are sudden extreme changes in the relative humidity of 
the atmosphere. The conservation of the paint layer depends in the 
first instance on the condition of the supports, whether panel or 
canvas, and their reactions to the environment. 




Church furniture of painted wood was in common use in Italy 
from very early times, and an obvious development was the use of 
flat painted panels for wall decoration and for altar pieces. When 
large panels were required they had to be made by joining planks 
together, and for this purpose a cement was used made from hme 
and cheese which is, essentially, calcium caseinate. If the planks were 
thick enough, they were pegged as well with iron or wooden dowels. 
Very heavy panels were often additionally reinforced with cross- 
bars, either inserted wedge-fashion into a converging channel cut 
across the grain, or fixed in some other way to the back. Such rein- 
forcement was not possible in the case of thin panels especially when, 
as in the side leaves of a triptych, they were painted on both sides. 
Panels completely protected on both sides by paintings have proved 
to be very durable, whereas thin panels painted only on one side have 
been less able to withstand the influences that cause warping. 

The mechanical behaviour of painted panels is closely related to 
the environment to which they are exposed: 

1. When conditions of storage or exhibition are kept constant as 
regards temperature and relative humidity, a panel tends to attain a 
state of equihbrium and no warping, cracking, or movement of any 
kind need be anticipated. 

2. When a picture is transported to a locahty where the environ- 
mental conditions are different from those to which it has been 
accustomed, a certain amount of movement will take place. In a 
composite panel that is unprotected at the back, each constituent 
plank will tend to cup, and at this stage cracks may make their ap- 
pearance on the painted surface opposite the joints. 

3. In a picture gallery where the air is not conditioned, day-to- 
day variations in temperature and relative humidity are to be ex- 
pected, but normally these do not impose any great strain on the 
support. However, in addition, there are the long-term variations 
of a seasonal nature to be considered. When artificial heating is 
brought into action in the autumn and discontinued in the spring, 

B. After treatment 




the environmental changes may be considerable. These changes impose 
strain on the support and for this reason care must be taken especially in 
the spring and autumn to look for any signs of cracking and bUstering 
or weaknesses in the paint overlying the joints in composite panels. 

Temperature and humidity are the most important factors m the 
life of a panel paintmg. Exposure to extremes of temperature will 
cause desiccation and shorten the life of the picture, but extremes of 
humidity are the more potent and insidious cause of trouble, wood 
being a bad conductor of heat but responding quickly to damp and 
dryness. In the conservation of panel paintings, however, it is, above 
all, variation of relative humidity that has to be guarded against as 
this causes intermittent movements of the support that give rise to 
cracking and cleavage of the ground and paint layers. 

Easel paintings have been carried out on several other types of 
support — metals, slate, marble, &c. Of these the metals are the most 
important despite the fact that ground and paint layers do not adhere 
well to the non-porous surface. Metals are, moreover, good con- 
ductors of heat, and for this reason paintings on metal are more prone 
to damage by variations of temperature than by variations of the 
relative humidity of the environment. 

Panel paintings: mechanical damage and treatment 

I. Cracking. The most sensitive kind of easel painting is un- 
doubtedly the large wooden panel, especially when it is made up 
from thin planks and painted only on one side. Such a panel is prone 
to mechanical damage. It may be fractured even by Hfting or moving 
it carelessly. A less obvious cause of trouble may be the position of 
the picture on the wall, where it is perhaps exposed to a mild draught 
of air from a door or window, and the changing humidity will 
ultimately cause it to crack. This tendency to cracking explains why 
so many panel paintings have been treated in the past with a cradling 
or parquetry reinforcement at the back. The cradling has been apphed 
with the twofold object of giving mechanical strength to the panel 
and of confining its movements to one plane so as to prevent twisting 
or warping. Cradling consists of a series of wooden slats (‘stretchers’) 

B 6157 




fixed rigidly at intervals in the same direction as the constituent 
planks, i.e. along the grain of the panel; these stretchers have slots to 
accommodate another series of wooden slats {‘runners’) arranged at 
intervals at right angles to the grain and in contact with the panel. 
While the stretchers are fixed, the runners should have freedom of 
movement to allow for slight expansion or contraction of the panel, 
but so often the runners themselves become warped and immobilized, 
with the result that the cradling becomes the prime mover, and 
the panel has to adjust itself as best it may. For this reason, if a cradling 
has to be applied to strengthen a picture, it should be kept as simple 
as possible. 

Other ways of reinforcing a panel are by inserting X-shaped 
wedges or inlays across the joints of planks in order to prevent their 
opening, or to trust to a series of ‘buttons’ glued across the joint to 
hold the crack. Buttons are palettes of wood about 4 in. by 2 in. by 
J in., the size varying with the conditions. A common type of 
reinforcement is that achieved by gluing textile to the back of the 
panel, and this is often found where a panel has been enlarged by 
the artist in the course of painting the picture. Rubens often varied 
the proportions of his smaller panels, and reinforced the backs in this 
way. When a thin panel has been enlarged it is seldom that the dif- 
ferent members are in harmony, and w'arping often occurs. Some of 
these enlarged panels are so warped that it is difficult to contain them 
in their frames. Not infrequently textiles were fixed on to the face 
of panels before applying the painting ground, and this is a sound 
techmque because it not only strengthens the panel, but at the same 
time provides a good key for the painting ground. 

When a panel painting is observed to develop a convexity on the 
painted side, or a series of convexities or undulations corresponding 
to its constituent members, this should be regarded as an emergency 
requiring immediate attention. The trouble may arise from one of 
two causes: 

Firstly , it may be due to the fact that the picture has recently been 
hung in a dry room or exposed to heat, e.g. by sunning or by being 
near a calorifier. 



Secondly, if a picture has attained a state of equihbrium under 
conditions that are rather damp, say in a large private house, and has 
been lent to an exhibition, its exposure in the drier conditions of the 
gallery may cause dehydration at the back, where unprotected by 
paint, with consequent cupping of the panel (see Wood, p. 118). 

In both cases the picture should be taken at once to a cool basement 
room where conditions are known to be damper and laid flat, face 
downwards. The picture may usually be left in its frame, but any 
nails or metal cleats aroimd the edges that might restrain the natural 
tendency of the panel to revert to its original shape should be re- 
moved. A restorer with experience of panel work should then be 
consulted. He may consider that it will be sufficient to leave the 
panel to recover over a period of time, or that some special form 
of treatment is required. Recovery, which in the natural course 
of events may take many months, can be assisted by the judicious 
apphcation of pads of damp blotting-paper to the unpainted side. 

Afrer the panel has recovered its shape and been given any neces- 
sary reinforcement, cracks may remain on the pamted side, and these 
will require to be filled with stopping. Various materials may be 
used for stopping, one of the commonest being a putty made from 
plaster of Paris and glue to which has been added a httle stand oil. 
After the crack in the painting ground has been filled in, the surface 
of the stopping is painted in the colour and texture of the original, 
taking care that the new paint is level with the adjacent surface (see 
Inpainting, p. 175). 

When a panel painting is being sent from its normal environment 
to an exhibition, it is a wise precaution to protect it on the back with 
a material that is impervious to moisture, e.g. Polythene sheet, fixed 
with Polythene tape. Protection of the back in this way tends to 
counteract the effects of changes in humidity and warping is pre- 
vented. This precaution is only necessary when the panel is removed 
from the environment to which it has become accustomed. On its 
return home the sheet should be detached as there is no advantage in 
covering the back when the panel is known to be stable in its home 



2. Cleavage and flaking paint. Cleavage may take place between 
any of the adjacent strata of the painting. When it occurs in isolated 
areas it will commonly result either in the appearance of a bHster or 
in paint becoming loose on the surface. 

When cleavage is widespread, the condition is more serious. Ex- 
tensive cleavage is an indication that the panel is no longer able to 
fulfil its fimction as a support, and consideration has then to be given 
to the possihihty of its replacement, either by a fresh panel, or by 
some structural material that may serve the purpose more suitably. 
This type of operation, known as ‘transfer’, is one of the greatest 
dehcacy, and is only undertaken as an extreme measure by an 

When loose paint is found on the surface of a picture, action must 
be taken at once to prevent it falhng off. If movement of the picture 
is hkely to residt in loss of paint, emergency measures should be taken 
while the picture is stiU on the wall by sticking mulberry tissue over 
the damaged areas using as the adhesive dilute gelatine. If the glass 
cannot be detached from the front, the picture wiU have to be taken 
down carefully and removed from its frame. It is then laid face 
upwards in preparation for treatment. The cause of flaking may be 
due to the breakdown of an old repaint or to tension in the panel 
resulting from an advancing crack in the wood, but there is no doubt 
that, fundamentally, such instabUity is caused by exposure to uncon- 
trolled humidity changes (see p. 4). 

The flaking paint is reattached to the picture by injecting an 
adhesive under the raised paint — usually either a dilute solution of 
gelatine or a mixture composed of beeswax and resin — and applying 
gentle pressure and heat so as to flatten the paint. This work must be 
carried out with dexterity as the glue so quickly becomes tacky and 
the wax-resin adhesive shows a tendency to set prematurely. Opera- 
tions can be greatly facflitated by using the appropriate tools. Fine 
brushes and spatulas are required, and an electrically heated steel 
spatula having a gently curved blade is an invaluable instrument in 
preventing the adhesive from setting too soon. The degree of heat 
should be controlled within narrow limits so that the instrament can 



be employed to soften and flatten the paint without changing the 
shape of the fractured edges. 

It is important to draw attention to the danger of fing ering a 
blister before treatment, as the paint is brittle and under compression, 
and it may easily fly to pieces at a touch. 


The advantages of a textile support for easel painting came to be 
recognized in the fifteenth century with the discovery of the pos- 
sibflities of oil painting, and from this time onwards canvas was 
destined to become the principal form of painting support. It had 
the merit of hghmess and flexibflity. A pamting on canvas could be 
carried as a baimer and, within limits, it could be rolled, so that the 
large easel painting was no longer static in the sense of being an 
elaborate piece of furniture that must remain where it had been 

Canvas is the fabric in general use today for easel painting. It is 
woven from spun threads of cotton, hnen, or hemp, in tabby or twill 
weaves, and the fibres are protected from direct contact with the oil 
by a sizing of glue. The canvas is then primed in such a manner as to 
preserve the texture by covering it with a thin ground of linseed oil 
and white lead. This painting grotmd provides a uniform surface to 
work upon and one that will reflect hght, so that the artist can exploit 
to the full the translucent quahty of his paints. The canvas is stretched 
m a wooden frame provided with wedges in the comers so that it can 
be tightened or slackened at will. 

A stretched raw cloth canvas tightens when damp and expands 
when dry, but the effect of glue sizing and of priming and painting 
reduces such movements to a minimum, and may even cause them 
to operate in the reverse direction, some expansion taking place in 
damp conditions and some tightening in dry conditions. Very large 
thin canvases are those most hable to sag in their stretchers and, if 
allowed to do so, they may easily take on a permanent warp, especi- 
ally noticeable towards the comers, where undulations appear that 
soon collect dust. It is not a good thing, however, to tamper much 



with the stretcher keys after they have once been adjusted to a reason- 
able mean. It may be that the site occupied by the picture is exposing 
it to too high a temperature or relative humidity. The picture should 
be rehimg elsewhere for a year and kept under observation. Should 
there be no improvement it may be necessary to have the canvas 
removed from its stretcher and replaced in contact with a stretched 
raw cloth canvas. The additional canvas at the back will give sup- 
port to the picture and slow down its response to atmospheric 
changes in humidity. 

It is clear that in the case of canvas, the environment plays a part 
in conservation second only to its influence on panel. While the oil 
ground and the superimposed paint layers on a canvas are more 
adaptable to movements of the support than the gesso ground and 
paint layers on wood, the elasticity of the oil is gradually diminished 
with the passage of time; the paint layer becomes brittle and is easily 
cracked by movements of the support, due to humidity changes. 

An accidental knock at the back of an old canvas may cause local 
stretching, and result in the formation of a series of concentric 
cracks in the groimd and paint layer, a disfigurement not necessarily 
very serious, unless accompanied by cleavage and bhstering of the 

It is characteristic of canvas that it tenders with age, and gradually 
becomes weakened so that it is easily punctured or tom. Tendering 
may he due to a variety of causes but a major factor is certainly the 
cumulative action of sulphur dioxide absorbed from the atmosphere 
and its conversion to sulphuric acid by the catalytic action of iron.^ 
A canvas may become so rotten with age as no longer to be able to 
bear the strain where it is held to the stretcher by tacks. In this case it 
may be sufficient if the edges only are strengthened, and this is done 
by a process known as strip-lining. Eventually there will come a time 
when the whole fabric gets so frail that it can no longer he regarded 
as a satisfactory support, and then consideration has to he given to the 
possibility of backing it with fresh canvas, a process known as lining 
or, more generally, relining. 

* In tacking a canvas to a stretcher copper tacks should always be used. 



Canvas paintings: mechanical damage and treatment 

1 . Patching holes. Holes in canvas are repaired by backing the canvas 
with a patch, and fillin g either with stopping or with an insertion 
depending on the size of the hole. If the canvas is puckered it will 
have to be relaxed with moisture and flattened, and it may be neces- 
sary to remove hard paint from loose threads with a paint remover in 
preparation for the repair. In the case of a small hole the picture is 
laid face down on glass with the damaged area in contact with an 
oiled paper. A patch is prepared from a piece of raw cloth canvas 
larger than the hole, and it should be ‘chamfered’ by pulling several 
threads from the weave all round, thus thinning the edges so that 
the shape of the patch will not appear on the front of the painting. 
A suitable adhesive is a mixture of warm beeswax and resin, the patch 
being fixed with a warm iron; or a commercial rubber-paste cement* 
may be used or a glue-paste mixture. The picture is then turned over, 
the oiled paper removed, and the hole stopped with putty made from 
whiting and linseed oil in preparation for inpamting. 

If the hole is large, it is filled in by inserting a piece of primed 
canvas of the same weight and weave as the original canvas. The 
weave of the inserted piece is ahgned with the threads of the canvas, 
the contour traced, and the piece cut to the exact size and shape of the 
hole; it is then fitted in position, and secured temporarily on the front 
by Sellotape. A patch is appHed to the back as already described, and 
stopping is used on the front if required before inpainting. 

2. Mending tears. When tears in the canvas have to be mended, the 
tom edges are first brought together working from the back so that 
the weave can be distinctly seen, the broken threads being fined up, 
and the tear secured with Sellotape. The picture is turned over and 
the tear secured on the front. Then working from the back again, the 
Sellotape is removed and a patch applied as described above. In 
applying a long and awkward patch with a glue-paste adhesive, care 
must be taken not to stretch the material too tighdy as on drymg 
it may shrink and pucker the canvas. In using aqueous adhesives 
puckering of the joint due to differential shrinkage is a frequent 

* c.g. Grade 753 : National Adhesives Ltd., Slough, Bucks. 

i68 organic materials 

cause of embarrassment. Polyvinylacetate emulsion^ or epoxy resins^ 
are less liable to cause trouUe in this respect; they have the ment 
that they are used cold, and no great pressure is necessary in order to 
make good the joint. Wax-resin adhesives may also be used, and in 
this case the patch is apphed, as mentioned above, with a warm iron. 

3 . Strip-lining. When the canvas support is in good condition gener- 
ally, and only the edges are weakened where they have been attached 
to the stretcher, these can be strengthened by applying strips of 
canvas to the back of the painting. Raw cloth canvas is used of the 
same weight and texture as the picture. The strips are cut to fit, 
allowing sHghtly more material at the outside edges for fitting over 
the stretcher; the inside edges must not extend beyond the -width of 
the frame, and these edges are ‘chamfered’ as in preparing a patch. 
A strong adhesive is necessary for strip-lining, and the glue-paste 
mixture may be used, but the best cement for this purpose is probably 
Araldite loi as this does not shrink on setting, and there should be 
no tendency for the canvas to cockle. Should its removal ever be 
necessary, however, this could not be accomphshed by heat or solvent 
action, but only by rubbing down from the back with glass-paper. 
Whatever adhesive is used, the repaired canvas must be put in a 
press or kept under weights until the joints are secure. 

4. Relining. Relining is for canvas paintings what transfer is for 
panels: it is a method that can be used for saving a picture that is 
threatened by the failure of its support. Most old paintings of any 
importance have already been rehned. 

In relining, the painted surface is first reinforced -with tissue paper 
apphed with an aqueous adhesive; then, any pre-vious relining canvas 
is removed. If a painting is being relined for the first time, the old 
canvas is smoothed down before a new canvas is apphed. Relining is 
a major studio operation, but it is only by virtue of the fact that it 
can be expertly accomphshed that so many Old Master paintings 
survive in such excellent condition today. This work is never attemp- 
ted unless it is deemed to be essential for the survival of the picture 

* e.g. Texibond V4N: Messrs. Scott Bader & Co. Ltd., 109 Kingsway, W.C. 2. 

^ e.g. Araldite loi; Aero Research Ltd., Hinxton, Duxford, Cambridge. 



and then it is only imdertaken by a qualified expert. Apart from the 
general risks involved in relining, the chief danger is that character- 
istics of the artist’s technique may be impaired; for example the 
impasto may be flattened and a tonabty introduced that is foreign 
to the -work, but in recent years the technique of relining has been 
improved and brought more under control. When heat is required, 
it is apphed either by electric irons thermostatically regulated or by 
the use of a hot table, ^ and much ingenuity has been devoted to 
evolving methods of protecting the impasto during reli ning . 

The glue-paste adhesives that were originally in universal use have 
largely been replaced by various wax-resin mixtures containing 
inert materials such as China clay, powdered chalk, &c. Wax rehning 
has several advantages: the wax consohdates the paint layer by 
impregnation through the cracks from back to front, it avoids the 
compHcarions that may result from using aqueous cements, and it 
can be easily removed at any time without strain when further 
rehning is required. It is possible to carry out rehning without heat- 
ing by using a cold-setting emulsion adhesive such as Texibond 
V4N and this has the advantage over glue that it remains permanently 
flexible. It is only apphcable, however, in cases where the paint layer 
is in such good condition that it does not require to be consohdated 
by impregnation with wax. 

In the preceding paragraphs treatment has been prescribed for 
dealing with certain types of deterioration arising primarily from 
failure of the panel or canvas support. It falls now to consider in how 
far the paint and varnish layers may suffer damage in themselves, or 
from what may be regarded as surface agencies. 


The paint layer is composed of insoluble particles of pigments 
mixed together in the greatest complexity, each particle being sur- 
rounded by an impervious film of medium. In egg-tempera and oils 

' Ruhemann, H., ‘The Impregnation and Lining of Paintings on a Hot Table’, Studies 
in Conservation, 1953, I, p. 73. 



the medium isolates the pigments from moisture and noxious gases. 
This is not so in the case of water-colours or gouache, in both of 
which gum Arabic is the essential binding medium. White lead, 
which in water-colour is so readily blackened by the action of 
sulphuretted hydrogen in the atmosphere (p. 80), is immune when 
ground in oil or egg-tempera. 

Pigments that are chemically incompatible in water-colour may 
often be mixed in oils with much less chance of their interacting with 
each other, and when resin is added to the oil it is even possible to 
use colours made from mixtures of pigments containing sulphur and 
copper without fear of subsequent chemical interaction which would 
lead to loss of briUiance. This has always been well known, and El 
Greco, for example, mixed such potentially incompatible pigments 
freely by employing isolating resin varnishes. It would seem therefore 
that in easel paintings there is no problem as regards the preservation 
of the pigment constituent of the paint layer. Whenever chemical 
changes have occurred, they are usually irreversible. Two of the 
commonest examples to be found on Old Master paintings are 
associated with ultramarine and copper resinate green. 

Ultramarine is occasionally found to have lost its briUiance and to 
have become mottled white in appearance. This is due to the action 
of acid, probably arising as a legacy from some iU-advised process of 
cleaning in the past. In water-colours this so-caUed ‘ultramarine 
sickness has even been traced to acidity in the paper support. 

In many early paintings it has been observed that areas that were 
originally a brilhant green have now assumed a dirty brown colour. 
This is due to a chemical change which copper resinate greens under- 
go, particularly in alkahne conditions. Since alkahne reagents (e.g. 
soap, wood, ash, ammonia) were used in the early days for cleaning 
paintings, these may be responsible for the browning that is so often 
observed. Alkahs can also destroy the colour of Prussian blue, a pig- 
ment vvddely used in the palette with yeUows in composing mixed 
greens, and in this case also there is a tendency for the greens to go 
brown and there is no possibihty of restoring the orig inal hue. 

It is questionable whether the amount of sulphurous gases present 



even in industrial atmospheres would be enough to affect pigments 
bound in oil or tempera, but it is a different matter when soot, which 
is greasy and contaminated with sulphuric acid, is allowed to accumu- 
late on a painting: it becomes ingrained in the impasto and craquelure 
and the weave of the canvas, and causes deterioration of all the materials 
with which it is in contact. Dust in any form is harmful and for this 
reason paintings that are of any value should never be hung against 
the chimney-breast of an open fire, or above a radiator; apart from the 
undesirable temperature and humidity conditions in these situations, 
the dirt problem is intensified by convection currents. The main reason 
for keeping pictures imder glass is to protect the paint layer from 
noxious gases and from dust more effectively than can be done by 
the varnish itself. When a picture gallery is air-conditioned glazing 
is unnecessary. 

As an oil painting matures there is a tendency for it to go down in 
tone; the medium becomes more translucent and darker in colour. 
The increased translucency is due to an increase in the refractive 
index of the oil on ageing, and this cannot be prevented. The main 
form of deterioration in the paint layer, however, is that caused by 
the darkening of the oil medium. Linseed oil is a complex mixture 
of fatty acid glycerides, certain of which (e.g. linolenic) tend to become 
dark in colour on ageing. When an oil painting is kept in a dull 
light the rate of darkening is intensified, and although the main cause 
of such darkening is usually due to the deterioration of the varmsh 
(see below) the change in the medium itself is a factor of importance. 
The pernicious practice of rubbing over the surface of a painting 
indiscriminately with linseed oil under the mistaken idea that this 
will help to preserve it merely aggravates the tendency to darkening 
and cannot be too strongly condemned. Darkened oil paint may be 
restored in large measure by exposing the picture to sunhght which 
bleaches the discoloured ingredients of the oil. The medium of 
tempera painting is in a different category. It has no tendency to 
darken or lose its optical quaUties, and as it has a comparatively low 
refractive index, delicate nuances of colour can be appreciated in this 
medium to a degree impossible in oil painting. 




The final coat of varnish is apphed to a picture to enhance its ap- 
pearance, to give depth and luminosity to the colour, and unity to 
the composition. At the same time it gives the picture some mechani- 
cal protection and shields it from direct contact with the atmosphere. 
But varnish is impermanent, and as it decays and is removed and 
replaced many times during the life of an Old Master painting, it is 
necessary that it should be of such a nature that it can be easily 
removed, even in the decayed condition, without exposing the paint 
layer to the action of solvents that might soften it, and without 
involving any mechanical strain that might weaken it. It is hardly 
necessary to say that a varnish must also be transparent and remain 
so, and it must be tough; and to be of maximum optical value it 
should have a high refractive index approximating to that of the 
dried oil films (ca. 1*53). Further quahties are required, but the few 
enumerated above are sufficient to restrict the field to what are called 
the spirit varnishes, i.e. those made from resins that are directly 
soluble in volatile solvents such as turpentine or white spirit, and that 
on evaporation of the solvent leave behind a film of resin having the 
desired quahties. Of the natural resins, mastic and dammar come 
nearest the mark, and in spite of certain deficiencies are in most 
general use today. Certain synthetic varnishes have been recently 
introduced and these are discussed in detail later (p. 178). 

Varnish deteriorates by blooming^ and by becoming brittle and 

[a) Bloom. This word has been used to describe the dull bluish- 
white appearance that develops in a transparent film and leads to a 
cloudiness. The precise cause of blooming is not known, and while 
many factors may be involved, damp is certainly a major factor in 
the blooming of varnish. Dammar and mastic are sensitive to moisture 
which affects the varmsh film after it is dry, causing it to become 
clouded with a bluish-white mist. The blooming of varnish may 
also be caused by the use of inferior materials or incorrect solvents 
in its formulation. 

’ Brommelle, N., Museums J. {1956), 55, pp. 263-6. 



If dealt with in the early stages, bloom can be removed from the 
surface of the picture by rubbing with a soft silk pad, either dry or 
charged with a httle wax polish, but if left unattended, the bloom 
tends to work its way into the film and it can then no longer be 
removed by surface treatment. 

In the case of pictures that are prone to bloom, a paUiative may be 
found in the use of a microcrystalhne wax pohsh (see Appendix XII). 
For the precautions necessary to prevent bloom forming during re- 
vamishing, see pp. 176-8. 

{h) Embrittlement and loss of transparency. So far as our present know- 
ledge goes, all spirit varnishes based on natiural resins become brittle, 
disintegrate with age, and develop a yellow colour. The yellowing is 
most pronounced in varnished pictures that are exposed to a bright 
hght. While dayhght prevents oil paint from yellowing it has the 
opposite effect on dammar and mastic varnishes. Dayhght is essential 
for the Hfe of a painting in the oil medium, but exposure for long 
periods to direct sunlight is not recommended as this causes pre- 
mature ageing of the varnish and hastens the day when its replace- 
ment win be necessary. 

In old pictures that may still retain a coating of oil varnish, e.g. an 
od-run copal, the decay of the varnish is accompanied by crazing 
which may communicate itself to the tmderlying paint and cause 
serious damage to the picture. 

Whatever the nature of the varnish the breakdown of the film 
exposes a much larger surface area for the condensation of moisture 
and accumulation of dirt, and when the varnish has become badly 
discoloured and porous and is no longer protective its removal is 


I. Removal of varnish 

Films of soft resin varnishes (mastic and dammar) are removed 
from a painting by applying solvent on swabs of cotton-wool to 
small areas at a time, the dissolved varnish being absorbed by the 
wool and the swabs changed frequently during the course of the 



work. The choice of solvent is crucial as it must dissolve the varnish 
without affecting the underlying paint. Many solvents for resin are 
available — alcohol and acetone are the two most generally used — 
but they must be diluted with a restrainer (turpentine or rectified 
petroleum) in such proportions as to ensure that the hquid has no 
action on the paint lay^er, but dissolves the varnish without undue 
friction. In order to adjust the proportions of solvent and restrainer, 
cleaning tests are first made on the edge of the picture. It may be 
found that a mixture of alcohol and turpentine in the proportion of 
4 vols. to 20 vols. removes the varnish without affecting the paint; 
a 2/20 mixture may be equally satisfactory, whereas a 1/20 mixture 
may be found to be effective only on continued rubbing and this is 
undesirable as it would be likely to damage impasto. In such a case 
the 2/20 mixture might seem to be the best to employ. This would 
then be tested on varnish overlying small areas of bright colour, e.g. 
vermihon (which is usually most sensitive), using for the purpose of 
the test a wisp of cotton-wool twisted round a match-stick. The swab 
is closely examined in order to discover whether, as a result of clean- 
ing, any pink colour has been communicated to the cotton-wool. If 
so, the cleaning solution is too strong and it must be further diluted 
with turpentine. When the mixture has passed these tests satisfac- 
torily, parts of the painting are selected and rectangular areas cleaned 
of varnish in order to form an estimate of the condition of the under- 
lying paint and the change of tonahty that may be anticipated as a 
result of cleaning. This prehminary routine is a necessary prelude to 
the cleaning of any important picture. 

In stripping a picture the restrainer must always be readily acces- 
sible. In practice, the operator can hold the cleaning swab in one 
hand and the restraining swab in the other, so that should there be 
any indication of the softening of the paint, as by the appearance of 
a trace of colour other than brown on the cleaning swab, he can flood 
the area with restrainer and arrest the action. The areas are finally 
wiped over with the restrainer to remove the last of the solvent and 
any softened residues of varnish that may remain on the picture. In 
the case of high impasto or resolute brush-work, such as is found in 



paintings by Rembrandt, hollows may be filled with dark varnish 
while the raised portions are hght and possibly abraded. The cleaning 
of such paintings requires time and patience as the work may have 
to be carried out very largely by using small wisps of cotton-wool 
on match-sticks as described above, rather than larger swabs held in 
the fingers. 

The experienced restorer will be able to estimate beforehand what 
the paint layer can stand in the way of cleaning. He will recognize 
the existence of any soft resins used by the artist in the course of 
painting the picture, and ensure that such a painting is not damaged 
by his cleaning solutions. He will consult with the curator regarding 
the removal of any repainting that dates from a former restoration. 
In cleaning a picture it is often found that previous repairs have been 
concealed by unnecessarily large areas of repainting, and the removal 
of such restorations, followed by inpainting of the stopping only, 
win usually be desirable. 

2. Inpainting 

The term ‘inpainting’ is used in picture restoration to emphasize 
the fact that the original paint is sacrosanct and must not be covered 
with modem additions of colour. The only exception is where the 
picture is abraded and part of the design rubbed away; here it may 
be necessary in the interests of aesthetics to retouch the paint layer 
with colour — a dehcate operation for which paints are employed 
that can readily be removed if necessary . Where inpainting is re- 
quired, either in oil or tempera paintings, the medium that is chosen 
is egg-tempera. This is apphed irrespective of the original techmquc 
because it is not subject either to darkening or to increase in trans- 
parency; the oil medium, as has been mentioned, is subject to both. 
Tempera has the one disadvantage that it becomes insoluble, but this 
is no serious inconvenience when it is apphed to areas of stopping 
only. In case it should ever be necessary to remove the modern paint, 
it may be apphed over an isolating varnish of dammar in turpentine 
and this is the technique generahy adopted, each coating of tempera 
being given a thin coating of dammar varnish before the next is 



applied. As the dammar can always be softened by the use of resin sol- 
vents, this makes it possible to remove the retouchings at any time. 

In order that the quahty of the inparnting should be in harmony 
with the rest of the picture, it is necessary that the transition from 
stopping to finished painting should follow the same course as that 
adopted by the artist. Thus, the stopping should be coloured to 
match the ground of the painting; if drawing or under-modelling 
is present, this should be reproduced on the stopping, and the under- 
painting and glazing should be bmlt up as nearly as possible in the 
manner adopted by the artist himself. The inpainting should be 
carried out using stable powder pigments ground in the medium and it 
is important that the finished restoration should be of shghtly hghter 
hue than the adjacent passages of paint, a dark retouching being much 
more obtrusive than a Hght one. The aim should be not to deceive; 
it should be to conceal forms of damage that would distract the 
attention and by their presence make it impossible to appreciate the 
picture as the artist intended. A trained eye can usually detect restora- 
tions but, should the matter be in doubt, suspicions can often be 
confirmed by inspection of the painting in a darkened room under 
ultra-violet radiation, or by infra-red photography, or X-radio- 

After the painting is dry the picture is ready for revamishing. 

3. Revarnishing 

As aheady mentioned, mastic and dammar varnish are the picture 
varnishes in most common use. It is difficult to decide which is the 
better as the resins themselves are complex mixtures^ and subject to 
much variation in quality. Mastic seems to give the more brilhant 
fimsh, but it requires a turpentine solvent which may be one reason 
why it becomes so yellow on ageing. Its use was abandoned by the 
National Gallery in 1949 in favour of dammar vamish,^ which, in 

' Mills, J. S., and Wemer, A. E. A., ‘Partition Ciuromatography in the Examination 
of Natural Resins, Journal of the Oil and Colour Chemists Assoc. 1954, 37, p. 131. 

^ Lucas, A., and Brommelle, N., ‘Failure of Synthetic Materials in Picture Conserva- 
tion’, Museums Journal, 1953, 53, p. 149. 

23 - (;iLr jiRONzi; nuAtJON ouNAMiiNr ihom 

A. As excavated 

B. Aker restoration 


. 3000 B. C.) 



contrast to mastic, may be dissolved in white spirit and is believed 
to be less prone to yellowing. 

Dammar picture varnish is made from selected dammar resin by 
dissolving it in white spirit without the aid of heat. The resin is sus- 
pended in a gauze bag in the solvent contained in a covered jar, and 
the hquid shaken or stirred at intervals over a period of some weeks. 
A suitable concentration would be about 4 oz. of dammar in i pint 
of white spirit. A htde stand oil is added, finally, to act as a plasticizer 
to toughen the frhn and to help brushing-out quaUties, but never in 
excess of 5 per cent., i.e. i fluid ounce to i pint of varnish. 

A picture should never be varnished until the paint is quite dry. 
The surface must he free from dirt, grease, and wax, and to ensure 
this it is rubbed over tightly with white spirit and allowed to dry 
before beginning operations. The varnish may then be applied by 
brush or spray. The work should be carried out in a warm dry room 
free from dust and draughts; care should he taken that the varnish 
is not chilled but is several degrees above the temperature of the 

If a brush is used, it should be a special varnish brush made with 
bristles of the finest quality and kept exclusively for the purpose. 
The varnish is applied to the picture thinly in a series of overlapping 
squares, say 8 in. by 8 in. (depending on the size of the picture), using 
first horizontal strokes and then immediately afterwards vertical 
strokes. In this way any excess of varnish is removed and a film is 
left of uniform thickness free from brush marks. A second coat may 
be applied if considered necessary after the first has dried, but the aim 
should be to apply the thinnest varnish film consistent with realizing 
the desired optical effect. 

When the spray is used — and spraying has much to commend it 
especially when there is high impasto — the fan-shaped type of nozzle 
is best. The pressure and the distance of the jet from the picture have 
a bearing on success and will be determined to a large extent by the 
viscosity of the varnish. In using the spray the jet must not be too 
near the painting, because it will then deliver varnish that is too 
liquid to control; on the other hand, if it is too far away, it will tend 

B 6157 




to give a matt effect by covering the surface with minute spots or 
scales of resin that are too dry to coalesce, thus affording no protection. 

In theory, the thinnest uniform coating should be obtainable by 
spraying, but in practice a uniform coating is difficult to achieve 
without budding up a film that is at least comparable in thickness 
with that obtained by brushing. 

When a picture has been varnished it is allowed to remain on the 
easel till the varnish is dry; it should then be kept under glass for a 
few weeks until the varnish becomes really hard, when there will be 
less tendency for the surface to bloom. 

Synthetic resin varnishes. Various types of synthetic resin are now 
available and research is constantly proceeding with the aim of pro- 
ducing a synthetic varnish suitable for paintings that will have all of the 
required characteristics (p. 172) and none of the demerits of natural 

One of the first synthetic resins to be tested was polyvinyl acetate 
which has been used for many years on quite an extensive scale in the 
U.S.A. The choice of solvent is important as there are many solvents 
for this substance that are too active for appHcation to oil paintings. 
Alcohol only should be used for preparing a varnish for brushing 
purposes and the standard solution is made by dissolving 20 grammes 
of polyvinyl acetate in 100 ml. of ethyl alcohoh without the apphca- 
tion of heat. This may take a long time but solution is facilitated by 
stirring and shaking. The standard solution may be diluted, if neces- 
sary, to half or quarter strength, by adding ethyl alcohol. This 
formula is given by Bradley^ together with a rehable spray formula 
for a polyvinyl acetate picture varnish composed as follows: 

Standard solution of polyvinyl acetate in ethyl alcohol 

(20 per cent, w/v) 250 ml. 

Ethyl alcohol 250 ml. 

Cellosolve acetate (or ceUosolve) ..... 100 ml. 

Diacetone alcohol . . . . . . . .35 ml. 

* Of strength 96 per cent. It should be noted that polyvinyl acetate does not dissolve 
in absolute alcohol. 

Bradley, M. C., jr., The Treatment of Pictures, Cambridge, U.S.A., Art Technology, 




Attention has also been devoted to the esters of polymethacryhc 
acid, which have the advantage that they are soluble in hydrocarbon 

These two types of resins are linear polymers of relatively high 
molecular weight. Hence, for a given soHds content, the viscosity of 
their solutions is high and this gives rise to practical diflBculties in their 
application. These can be overcome by using a varnish based on 
polycyclohexanone^ which is a polymer of low molecular weight 
similar in physical properties to the natural resins. This is made up 
in white spirit and yields a relatively non-yeUowing fihn which, un- 
plasticized, is rather glossy and rather brittle. The glossy nature of the 
film is no disadvantage as the optical quality can easily be adjusted 
after the varnish has dried, by the apphcation of wax poHsh (see 
below). The brittleness, however, is a problem that is still engaging 
the attention of manufacturers, and several formulations are already 
on the market. It seems certain that with the discovery of a suitable 
plasticizer, this varnish will become a serious competitor to varnishes 
of the conventional type. 

4. JVax varnishing 

Wax varnishes are merely polishes, because even if appHed thinly 
to the painting by brush, the wax residue that remains on the picture 
after the solvent has evaporated is pohshed by using a soft long-brisded 
brush. The resulting gloss may be enhanced, if desired, by further 
polishing with a soft cloth. The surface of a wax varnish even when 
pohshed is duller than that given by resin or synthetic varnishes, and 
wax may be apphed over these to reduce their high gloss. Resin and 
synthetic varnishes are never apphed over wax. 

A common formula for wax varnish contains 4 oz. of white 
beeswax dissolved in one pint of distilled turpentine with the aid of 
a water-bath maintained at about 60° C. An improved form due to 
A. E. A. Werner is the wax salve made from microcrystaUine 
parafiin wax and Polythene wax (see Appendix XII). This leaves 

^ This material is available commercially as Resin AW2, manufactured by Badische 
Anihn u. Soda Fabrik, Ludwigshaven; an English equivalent, Resin MS2, is manufactured 
by Howards of Ilford Ltd., Ilford, Essex, 



a surface that is relatively non-sticky and the salve may be used either 
as a pohsh or, in the case of smooth painting, for ehminating in- 
grained dirt. 

In order to wax-pohsh a picture, the surface is first rubbed very 
gently with a piece of soft silk to remove any dust. Some of the salve 
is then placed in the pahn of the hand, where it quickly softens. A 
piece of cotton-wool wrapped in thin silk is charged with the soft- 
ened wax, and used to pohsh the picture with the minimum of pressure. 
In the case of a picture ingrained with dirt, the pohsh is apphed direct, 
using the cotton-wool without the covering rag. The waxed cotton- 
wool collects the dirt, and finally the surface may be polished (using 
a new covered pad) by the method described initially. Large canvases 
may require some support from the back (by a pad of papers or the 
like) during cleaning and polis hing 


1. Insect attack on panels. Painted panels are sometimes found to be 
attacked by wood beetle (Anobium), the larvae of which tunnel into 
the wood to emerge eventually from the surface as fuUy developed 
insects. Activity may be recognized by the appearance of holes or 
frass and, if action is not taken in time, the support may become so 
weakened that cleavage of the paint results. 

When insect pests are found to be active, the painting should be at 
once isolated to prevent adjacent pictures becoming infected. It must 
then be sterihzed, using an insecticide that will not damage either 
paint or varmsh, the most satisfactory form of treatment being fumi- 
gation with carbon disulphide (see p. 124). To confirm that steriliza- 
tion has been successful, the panel should be inspected at intervals of 
six months after funugation. If the old insect-holes are filled with 
wax at the time of fumigation, the appearance of any new ones will 
be readily detected. 

2. Fungoid attack on canvas. A canvas that has been relined with a 
glue adhesive is prone, under conditions of damp heat, to be attacked 
by moulds. The appearance of fluffy white growths, visible especi- 


ally on dark passages of paint, is an indication of fungoid activity, 
and if observed through a reading-glass will likely be found growing 
on the glue where it is exposed in the craquelure. To arrest and pre- 
vent further growth, the surface of the painting should be rubbed 
over hghtly with soft pads of cotton-wool and the picture exposed 
to sun and air. This may be all that is required. The glass, however, 
should be sterilized, before it is replaced, by cleaning with cotton- 
wool that has been moistened with formalin. In very bad cases it 
may be necessary to use formalin on the actual painting, but this 
should be left to the discretion of the restorer who will know from 
his experience or can discover by testing whether the picture is in a 
condition to be treated with aqueous solutions. The back of the 
canvas must also be considered and may be given permanent pro- 
tection against mould growth by the use of a non-volatile fungicide 
such as Santobrite (p. 28). 

For Further Reading 

Constable, W. G., The Painter’s Workshop. Oxford University Press, 1954. 

Mayer, R., The Artist’s Handbook of Materials and Techniques. Faber and Faber, n.d. 
Stout, G. L., The Care of Pictures. Columbia University Press, 1948. 





The conservation of museum objects of metal is a study in itself. 
Metals form a heterogeneous though well-defined group of materials, 
almost all subject to corrosion — that is, to the loss of metallic pro- 
perties with the formation of mineral incrustations. This is due to a 
series of chemical or electro-chemical reactions, and disintegration 
may be slow or accelerated depending on the nature of the metal and 
the conditions to which it is exposed. That metallic corrosion is 
accompanied by a change in appearance is fortunate, as this calls 
attention to the fact that chemical change is taking place, and the 
sooner the object is treated the better its chance of survival without 
loss of character. 

In order to arrest corrosion, it is necessary to discover and remove 
the cause. This may be, and often is, the presence of some active 
element such as chlorine that has entered into chemical combination 
with the metal, in which case it can only be removed by a process of 
counter-attack involving other chemical changes that are appUed in 
the laboratory under scientific control and that can be rehed upon to 
achieve this end. The effect of such treatment may be striking, as 
when an object concealed under a gross incrustation of minerals is so 
dealt with that its original metaUic quahty is restored; but such an 
extreme change in appearance is not the inevitable result of chemical 
treatment. It is often possible by the exercise of ingenuity to arrest 
corrosion without sacrificing the minerahzed or patinated surface, 
and there are many cases where this is desirable. The nature and 
condition of the specimen will determine what freedom of action 
exists and to what degree stabihty can be regained with the minimum 
of sacrifice. 



In dealing with corroded metals there are, in fact, three possible 
methods of treatment: the use of solvents, the use of chemical and 
electro-chemical reduction, and mechanical methods; and while it 
will be convenient to study these methods individually, it should be 
understood that they are not mutually exclusive, but are applied as 
occasion demands in combination. Thus, the use of solvents and re- 
duction methods are each dependent for ultimate success upon the 
apphcation of certain forms of mechanical treatment (brushing, 
washing, &c.) and there are many cases where the best results can be 
obtained only by a combination of all three processes. Metals have 
individual characteristics; they are susceptible to different chemical 
reagents, they corrode in different ways, and in doing so they form 
different kinds of incrustations that must be recognized and studied 
before it is possible to decide upon the best form of treatment. 
Chemical methods involving the use of specific solvents can only be 
described in dealing with the metals themselves. By contrast, the 
methods of reduction and the mechanical methods are of general 
apphcation to all corroding metal, and a description of these processes 
is introduced, therefore, as a preamble to the study of the individual 
metals of antiquity. 

In considering what treatment to apply, a physical examination is 
of the greatest importance. This is carried out with the aid of a lens 
and, if necessary, by exploring with a needle or in the case of ferrous 
metal by using a magnet. It is necessary to form some estimate of 
the thickness and regularity of the encrusted layer, strength of the 
residual metal, presence of ornament, taking special note of fine 
detail and of any cracks. To reveal the presence of hidden ornament 
in cases where the metal is heavily corroded, it may be necessary to 
use X-rays. Methods of treatment are based upon this preliminary 
assessment, and tables are given at the end of each chapter (save in 
the case of gold), to assist in deciding what may be the best pro- 
cedure. For the apphcation of the methods recommended reference 
must be made to the text, but whether the suggested methods are, in 
fact, the best to apply in the circumstances can only be decided by 
the knowledge, experience, and resource of the operator. 



If metals are considered as raw materials, it is noteworthy that few 
are found in nature in the free or uncombined state. They occur in 
combination with non-metaUic elements in the form of minerals, and 
these provide the ores from which the metals are won by smelting. 
Minerals, rather than metals, are thus the stable forms imder natural 
conditions, and it is to these minerals that metals tend to revert, 
especially when they are buried in the ground for prolonged periods 
of time. It has been observed that, other things being equal, the more 
easily a metal is won from its ores, the greater is its stability. Thus, 
metalhc tin which is obtained with the minimum expenditure of 
energy from the oxide (cassiterite) shows less tendency than copper 
to become reoxidized. Because of its relative stabdity, tin has been 
used as a coating for protecting copper vessels from at least the begin- 
ning of the Christian era, and even in this attenuated form it is often 
found to have survived in excellent condition. 

When metals are buried in the ground, the rate of corrosion is 
intensified according to the degree of acidity of the soil, its porosity, 
and the presence of naturally occurring soluble salts. These substances 
in the presence of moisture conduct electricity and are known as 
electrolytes. Metalhc corrosion is an electro-chemical phenomenon. 
The absolute strength of electric currents flowing over corroding 
metal can be measured, and by comparing such measurements with 
the rate of corrosion, change of weight. See., it has been shown that 
the whole compHcated cycle depends on electro-chemical principles. 

Metals that have been buried and have already suffered some degree 
of corrosion have a comparatively porous surface and are thus hable 
to retain traces of salts. Often these salts are sealed up in what ap- 
pears to be a stable incrustation or patina, but should they be exposed 
to moisture and oxygen they are likely to give rise to fresh activity, 
with consequent pitting of the surface and possibly serious dis- 
figurement. Metals behave differently, however, in their reactions to 
non-metals. Sometimes there is a progressive growth of the surface 
minerals which increase at the expense of the metalhc core; in others, 
the mineralization is compact and stable and, after prehminary de- 
velopment, it tends to inhibit further change. Where development 



has been slow and uniform the intricacies of shape and ornament are 
preserved and the appearance of the object may even be enhanced 
by the colour and texture of the patina. A patina that has developed 
slowly may sometimes be evidence of antiquity and a fine patina 
win determine the market value of an object. It is necessary, therefore, 
in considering the preservation of metal objects to consider also the 
preservation of their patinas when these may be taken as evidence of 
age or are in themselves of aesthetic value. The patinas of Chinese 
bronzes, for example, which have matured through the years in 
ground substantially free from soluble salts, have an irresistible appeal 
for the collector; but the danger is that mineral patinas which are 
apparently quite stable may become unstable if exposed to damp, and 
when corrosion becomes active it is sometimes a most difficult task 
to arrest it without serious loss of surface appearance. 

Burial in the earth, however, is by no means essential for the corro- 
sion of metals. Even when objects are kept indoors, moisture and 
oxygen are enough in themselves to cause dulling of the surface due 
to the formation of a layer of metaUic oxide. The results of attack by 
sulphurous gases that are present in industrial atmospheres are even 
more readily apparent, yielding the darker films of metallic sulphide 
known as tarnish; but disfigurations of oxide and sulphide are purely 
superficial and readily amenable to treatment. 

The individual metals of antiquity are few in number — gold, silver, 
copper, lead, tin, and iron, but they have been mixed together as 
alloys, intentionally or otherwise, from remote periods. Thus, elec- 
trum (an alloy of gold and silver) is much more common than pure 
gold, and ancient silver almost invariably contains copper. Copper 
and tin were alloyed intentionally to make the more useful metal 
called bronze, and lead was often added to the mixed metal. But there 
is evidence that tin was not always distinguished from lead by the 
early metallurgists; analyses sometimes reveal the presence of both 
or even a preponderance of lead when this could only yield an alloy 
that was less serviceable for the purpose intended. Some of the earHest 
examples of bronze from Ur of the Chaldees contain such a small 



quantity of tin that it is doubtful whether the composition was in- 
tentional and it may therefore be misleading to use the term bronze 
in this connexion. Lead and tin have themselves been alloyed together 
from remote periods. Copper and zinc were alloyed intentionally in 
Roman times for coinage, but here again, there are isolated examples 
of brass occurring in earHer civilizations (China), and to what extent 
fhis was intentional can only be determined by further study. 
The manufacture of steel is a comparatively modem process, but the 
carburization of iron was discovered many years previously at the 
armourers’ forge, where iron and slag were hammered in the pre- 
sence of carbon and wrought, possibly by chance, into steel, an alloy 
superior to all others for the manufacture of sword-blades. Such a 
chance occurrence might possibly provide a basis for the celebrated 
sword-blades of fable and legend ! 

MetaUic corrosion has been described as an electro-chemical pheno- 
menon. This action can be demonstrated where two dissimilar metals 
are in contact in the presence of a conducting solution of a salt the 
essentials of a simple electric cell; under these conditions the baser 
metal is corroded preferentially as long as the two are in electrical 
contact, the more noble metal surviving by what is called cathodic 
protection’. By conducting such experiments with different pairs of 
metals it is found that the metals can be arranged in order of their 
nobleness from gold downwards, through silver, copper, lead, tin, 
and iron, the baser metal always being sacrificed in the presence of an 
electrolyte where it is in contact with another metal higher up in the 
series.^ An interesting example of survival by cathodic protection is 
that of the frail but perfect silver lamp discovered at Ur of the 
Chaldees among a mass of copper lamps (Pis. 22 A and b). The fact 
that only the silver lamp survived was significant, and the condi- 
tion of the copper lamps was such as to confirm that they had been 
preferentially corroded in accordance with this theory. For the same 
reason, bright inlays of copper or its alloys are sometimes foun 

' It should be noted that when metals are alloyed, as in stainless steel, the posmon of 
the alloy in the electro-chemical series caimot be determined even approximate y. 



embedded in beavily rusted iron swords and knives, iron in relation 
to copper being the baser metal. 

It is not only the conjunction of two different metals that promotes 
electro-chemical decay. Where two or more metals are alloyed to- 
gether as tin and copper are in bron2e, the susceptibihty to corrosion 
is greater than in the case of an unalloyed metal and, when once it 
has set in, the decay of bronze proceeds more quickly than is the case 
with copper alone. In the same way base silver is corroded more in- 
tensely than pure silver. Indeed, it is no uncommon occurrence to 
fmd on treating an object that has all the appearance of being a cor- 
roded bronze, that it is in fact made of base silver, the baser consti- 
tuent (the copper) having corroded preferentially and covered the 
white metal with a heavy green deposit, thus concealing the true 
nature of the alloy. In such cases treatment can be apphed to remove 
the green incrustation, revealing the underlying silver enriched in 
appearance in proportion to the quantity of copper that has been 
leached from the alloy in the process of corrosion. 

From these considerations it follows that different metals are at 
different electric potentials relative to a surrounding electrolyte, and 
it is then but a step to the idea that if different potentials could be 
established on one and the same metal (e.g. by local inclusions in the 
surface, varying porosity and concentration of the electrolyte, &c.), 
an electric current would flow leading to an electrolytic form of de- 
composition. This has proved to be the case, and in some measure 
may serve to explain the more catastrophic forms of corrosion such 
as the rusting of iron and also the behaviom: of metals that have been 
subjected to the periodic action of saline water. 

There is, however, a brighter side to the picture, and it may be 
introduced by reference to a minor catastrophe which took place, not 
in the earth, but in a museum. A painted zinc label had been left on 
a corroded silver dish in a cupboard. One must assume that the cup- 
board was damp and that the corrosion layer on the silver was able 
to act as an electrolyte. After the lapse of some years it was discovered 
that the label had gradually subsided into the dish and the remains 
stood in a clear liquid to which part of the label had apparently de- 


composed. The liquid proved to be zinc chloride and on clearing it 
away it was found that the silver at the point of contact had been 
reduced from dull homy silver chloride to bright metal. Now, 
metallic zinc is a baser metal than any of the others to be preserved 
in museum work, and this simple reaction may actually be put into 
practice in modified form in the laboratory, for the treatment of 
metal objects that have been disfigured by corrosion. The method 
is known as electro-chemical reduction. 


I. Electro-chemical reduction 

The standard method involves the use of zinc and caustic soda, and 
the reaction is carried out in an iron or enamelled container with the 
aid of heat. Heat promotes chemical action, and thus enables results 
to be obtained reasonably quickly. 

Materials. The zinc is used in granulated form or in the form of a 
coarse powder, so as to present a large surface of contact with the 
object under treatment. It may also be used in the form of zinc wool 
and wrapped around the object. Zinc dust is not as a rule so effective, 
as it tends to clog together. 

Granulated zinc may be purchased as such, or zinc may easily be 
prepared in this form by heating the metal in an iron ladle (it melts 
at 419° C.), and pouring the molten metal from a height into a bucket 
of cold water. If required as powder, the granulated zinc is pulverized 
in an iron mortar to the requisite gram size, and it may then be passed 
through a sieve of about 30 mesh to the inch, which yields a granular 
powder sufficiently fine for the most dehcate operations. The advan- 
tage of using zinc wool is that wrapping ensures that electrical con- 
tact will be satisfactory; on the other hand, the object is concealed 
during treatment by the wool, and the progress of the reaction more 
difficult to follow. 

Commercial flake caustic soda is dissolved in water and employed 
as the electrolyte. The strength of the solution should be not less than 
10 per cent. (2 oz. to i pint of water). When in contact with excess of 



zinc this solution will be ‘spent’ after half-an-hour’s heating, i.e. 
evolution of hydrogen which is the potent factor in effecting reduc- 
tion will then have ceased. It is usual, therefore, to employ stronger 
solutions, say up to 20 per cent., for objects that are heavily corroded. 
It is better to change such solutions occasionally rather than to work 
with solutions in excess of 20 per cent, concentration. 

Method of procedure. Heavy objects such as celts, spear-heads, and 
the like are buried under a heap of granulated zinc in an iron basin, 
covered with the caustic soda solution, and the hquid boiled for an 
hour or so over a gas-ring or Bunsen burner, preferably in a fume 
cupboard as the vapours evolved are irritating. During this time, dis- 
tilled water is added, as required, to maintain the volume of the 
solution at the original level. Dehcate objects such as fibulae are 
buried in zinc powder, covered with a 10 per cent, solution of caustic 
soda, and heated over a steam bath for an hour, taking the same pre- 
cautions to maintain the volume of the solution. 

The result of such treatment is to soften the incrustation. In the 
case of bronzes, the green colour will disappear, and it will be found 
possible by brushing under a stream of water to remove much of the 
brown, muddy deposit, revealing a metallic surface upon which are 
retained any details or ornament. If the metaUic surface is not entirely 
clean, it will be necessary to repeat the reduction wdth fresh zinc and 
caustic soda. It is not permissible to leave vestiges of oxide un- 
reduced as they seal in traces of chloride, and the only satisfactory 
method of eradication is by further reduction followed by brushing 
and washing. If traces of chloride are allowed to remain, a further 
outbreak of corrosion is inevitable, sooner or later. In some cases the 
vestiges of oxide may be found to be covered by a film of metallic 
copper which will make them more difficult to remove. It is most 
important, however, that they be removed. This copper plating is 
not so hable to occur when the reduction is carried out in boiling 
solutions, as the plating is discouraged by the mechanical agitation. 

In cases where residual incrustations cannot be removed even with 
a steel or glass brush they will have to be picked or scraped off. This 
often requires the exercise of much patience, using a magnifier and 


• //‘i^ /’ 



The light areas arc tin itxiclc and tlic il.irk areas inainlv cuprite. The multi-layered structure 
indicates that the incrustation is td great age 



mounted needle, and going over the whole surface a small area at a 
time, picking off residual lumps of incrustation to expose the pockets 
of greenish-white pasty material (cuprous chloride) that underhe 
them. A subsidiary reduction is then required to ensure that salts 
exposed by this treatment are rendered soluble so that they can be 
washed out. 

Where equipment is available for electrolytic reduction it is con- 
venient to give the finished object half an hour’s treatment in the 
tank (Fig. 6, p. 194) as this facihtates the removal of the last traces 
of salt, thus cutting down the time required for washing. This 
operation must be conducted with thoroughness and is described in 
detail on pp. 197-9. 

In the course of reducing objects with zinc and caustic soda, the 
soda is spent and must be discarded, but a quantity of zinc remains, 
and, as it is rendered largely inactive due to the formation of a film 
of oxychloride and carbonate, it tends to accumulate in the labora- 
tory. It may be recovered for further use by washing it with water 
acidified with hydrochloric acid, then with distilled water, drying, 
and, if necessary, melting in the iron ladle, and regranulating. 

Variations in procedure. The electro-chemical reduction process as 
outlined above admits of several useful variations: — 

(i) If the object is encrusted with hme or chalk in addition to 
metalhc oxides, it may be that the reaction will proceed more regu- 
larly if the reduction is carried out using a 10 per cent, solution of 
sulphuric acid instead of caustic soda. 

(ii) Where local treatment only is required, as in exposing an 
inscription on the base of an Egyptian bronze statuette, reduction 
may be done by repeatedly applying a paste of zinc powder and 
sulphuric acid of 90 per cent, strength. Alternatively the paste may 
be made in situ on the surface of the object by applying a few drops 
of the acid and mixing in a small quantity of the zinc powder with 
a glass rod or a tuft of glass bristles held in a piece of glass tubing. 

(iii) Silver objects which are not heavily coated with horn silver 
respond well to reduction with zinc and hot formic acid, and even 
more readily when aluminium granules are used instead of zinc. 





The main advantages of electro-chemical reduction are that it re- 
quires no apparatus beyond what is readily available or can be easily 
improvised, and with the modifications described above it provides 
the means of reducing metal objects locally when desired. It is, how- 
ever, subject to the limitations that apply generally to aU methods of 
reduction. See p. 217. 

2. Electrolytic reduction 

As an alternative to the method aheady described, it is possible to 
effect reduction by using an electric current, see Fig. 6. The cor- 

6-/2yo/ts Supply 

roded metal object is made the negative electrode (cathode) in a 
suitable dectrolyte such as caustic soda solution, the positive electrode 
(ano e) eing of sheet iron; under these circumstances the reducing 
action is then dependent on the apphcation of an external e.m.f as, 
or exarnple, by the use of an electric battery. This method is some- 
times referred to as electrolysis. 

When current passes, hydrogen is evolved at the cathode with the 
res t t at the incrustation is gradually reduced and saline matter 
ecompose . As t e reduction progresses, chlorides are transferred 
from the cathode to the iron anode. This method can be very effec- 
ttve m cleamng badly encrusted metals of all kinds, but unless con- 
dittons are made to conform to accepted standards (as described 

below), comphcations m the form of secondary reactions may 
vitiate results. 



Equipment. The electric supply must be direct current, either from 
a secondary cell or accumulator, or it may be A.C. stepped down 
from the mains to about 12 volts and rectified (see Diagram, Appen- 
dix Xin). It is the amperage in the circuit that is the most important 
factor, however; this depends upon the resistance of the electrolyte 
and the size of the electrodes, but will vary during the process as the 
resistance of the incrustation on the object alters imder treatment. 

In practice, caustic soda of about 5 per cent, strength is chosen as 
the electrolyte, and two iron anodes are himg equidistantly, one on 
each side of the cathode, and only a few inches from it. Under such 
standardized conditions the rate of reduction will then depend 
mainly on the current density which can be regulated by a variable 
resistance to approximately 10 amps per square decimetre of cathode 
area, so as to give a steady vigorous evolution of gas. This figure is 
not critical for iron or steel. For copper and silver, however, the 
current density should not be allowed to fall below 2 amps per square 
decimetre, otherwise there is a tendency for a film of metal, usually 
salmon-pink copper, to be deposited on the surface of the object in 
the region of the chloride-containing incrustation, and this is difficult 
to remove. At first the resistance of the incrustation may be consider- 
able, but in the course of treatment, after perhaps an hour or more, a 
rise in amperage will indicate that the resistance of the incrustation 
has fallen, and this should be compensated by increasing the external 

Method oj procedure. A convenient arrangement is as follows: — A 
glass tank is filled to a suitable depth wdth 5 per cent, caustic soda 
and brass bars are laid across the top from which the electrodes are 
suspended by copper wires, thus allowing of easy adjustment. The 
object is attached to the negative pole of the electric supply, whilst 
the iron anodes are wired together and joined to the positive pole 
through an ammeter and adjustable resistance. 

The iron anodes are heavily attacked during electrolysis. They 
tend in time to deposit flakes of metalUc oxide, &c., in the solution, 
a state of affairs that may be prevented by tying them in nylon bags. 
But, even so, it is necessary to remove the anodes occasionally and 



scrub off the products of chloride attack. Graphite and carbon elec- 
trodes have also been used, but these break down in alkaline solutions 
and deposit a black fihn on the object. Attempts to brush this off only 
burnish it into the surface. An advantage of the iron anode is that 
when iron is plated on to the object, as often happens when prolonged 
electrolysis is necessary, it can be easily removed in the subsequent 
washing process. The chance of comphcations taking place is re- 
duced by using anodes of stainless steel. Nylon bags may be tied 
round the objects being treated if additional support should be re- 
quired, or if there is any chance of metal inlays, &c., f allin g off into 
the solution. 

Another refinement is to cover the surface of the electrolyte with 
floating glass capsules or short lengths of Polythene tube, called 
croffles,^ in order to minimize the escape of caustic spray which, 
though not excessive, tends to irritate the throat. The glass croifles 
are preferable to those of Polythene, and are used in the proportion 
of about 1 lb. of crofiles to 90 square inches of surface. 

Instead of using iron or steel anodes, the electrolytic tank itself may 
be of iron instead of glass, and made to serve as anode, but if origin- 
ally galvanized it wfll be necessary to remove aU traces of the galvan- 
izing metal which otherwise might plate on to the objects. This may 
be removed by running the tank as anode for a considerable period 
with a temporary iron (tin-plate) cathode, the caustic soda solution 
being maintained at the highest possible level in the tank until the 
last of the zinc has been stripped from the metal and the tank surface 
is clean. After washing and making up with fresh electrolyte, the tank 
is ready for use. 

In reducing a corroded object the duration of treatment will de- 
pend upon the nature of the incrustation, but its removal can be 
facihtated by taking out the specimen under treatment occasionally 
and brushing and scraping the surface. The object should be removed 
while the current is still running. In no circumstances should it be 
left in the tank after the current is switched off, otherwise it may 
become plated with any metaUic impurity present in the electrolyte. 

* Supplied by W. Canning & Co. Ltd., 77 St. John St., London, E.C. i. 



Caley ^ has referred to a comphcadon that is hable to interfere with 
the cleaning of bronzes that contain much lead. The lead is distri- 
buted in the alloy in the form of globules, and as these are gradually 
dissolved by caustic soda, the bronzes become pitted. Also the elec- 
trolyte tends to build up a concentration of lead plumbite. Lead may 
then be replated on the objects as a grey coating. Usually this is 
easily brushed off, but an electrolyte contaminated with lead might 
well plate this metal on to other specimens treated subsequently, with 
undesirable results. He recommends that, when necessary, the lead be 
removed from the electrolyte by plating it on to a large sheet of 
copper attached as a temporar)^ cathode to the cathode bar which 
normally carries the objects under treatment. The lead coating is 
easily stripped from the copper sheet later, by passing it through a 
bath of 10 per cent, nitric acid, and after washing in water it is ready 
to be used again should occasion require. 

This serves to draw attention to the importance of keeping the 
electrolyte clean. Lead can easily be detected in the electrolyte by 
withdrawing a small sample in a test-tube and adding dilute sulphuric 
acid, until it is shghtly acid. If a white precipitate of lead sulphate is 
formed, this is an indication that the electrolyte is contaminated with 
lead, in which case the lead may be removed by the aiudhary elec- 
trode method described above; alternatively a fresh lot of electrolyte 
may be made up to replace that which is contaminated. 


When an incrustation has been broken down by reduction there 
will remain on the surface of the metal a sludge of insoluble oxides 
and metalhc powder; this will also contain chlorides, rendered soluble 
by the treatment, as well as a residue of the electrolyte employed in 
the process. The sludge is removed by brushing the object under 
running water, and while this will remove most of the superficial 
soluble impurities at the same time, it is not by any means sufhcient 

’ Caley, E. R., Technical Studies, 1937, 6 , pp. 194-7. ‘Removal of Lead from the 
Electrolyte of Baths used for Cleaning Ancient Bronze and Copper Objects’. 



to deal with all. It will be readily understood that when an incrusta- 
tion has grown upon metal, the metal underlying the incrustation 
will be microporous, and when the incrustation is removed, the 
surface layers of the metal will behave as if they were a mass of 
capillary tubes, and these will retain a residue of chlorides. This cor- 
rosive matter is not easy to remove. It can be eliminated by prolonged 
soaking in changes of distilled water, or its removal may be speeded 
up by u sing hot water, but unless the capillaries can be flushed with 
the distilled water, there will always be the chance of chloride being 
retained in cracks or pits sufficient in concentration to cause a fresh 
outbreak of corrosion at a later date. 

Such considerations have been examined theoretically by R. M. 
Organ,! ^nd as a result of practical tests carried out under controlled 
conditions, using a conductivity meter, he has been able to recom- 
mend a form of intensive washing and also a particularly sensitive 
method of carrying out the silver nitrate test, i.e. the test that is used 
in practice to determine when objects are free from chloride. In this 
method it is recommended that the washing should be carried out 
with pure (distilled) water, by alternate heating and cooling, at le^t 
in the early stages, to ensure permeation. The heating causes any air 
or hquid in the capillaries to expand and partly leave the metal, and 
on cooling fresh water is sucked in. In dealing with a batch of bronzes, 
it is convenient to use a steam oven, in which are placed Pyrex 
beakers of distilled water containing the objects. The oven is kept at 
about 98° C. during the working day, the temperature falling over- 
night. This alternate heating and cooling ensures that the capillaries 
in the metal are flushed with water, and the progress of the washing 
is followed by testing the conductivity imtil it reaches a minimum. 

The procedure described above is of wide appHcarion to metals 
and many sihceous materials, notably to baked cuneiform tablets, 
but this method is not suitable where leaden objects are concerned. 
Hot distilled water reacts in a few hours with lead to form lead 
hydroxide which would cover the metal with a milky white film. 
A special procedure is required for washing leaden objects (see p. 260). 

! Organ, R. M., Museums Journal, 1955, 55, p. 112. 



The Standard test for chloride is carried out by adding silver nitrate 
solution in presence of nitric acid, when the formation of a white 
flocculent precipitate (of silver chloride) indicates that chloride is pre- 
sent in the solution that is being tested. 

In the final stages of washing, when testing is of the greatest signifi- 
cance, it is essential to be able to detect the presence of the merest 
traces of chloride remaining in the water. Organ has emphasized 
the importance of applying the test under optimum conditions. 
The volume of soak-water should not be much in excess of that 
required to cover the object and the water should be in contact with 
the object for some rime (until the maximum electrical conductivity 
of the water is attained) . Then the test is carried out under the follow- 
ing standard conditions which give the most sensitive results. About 
10 ml. of the water to be tested are poured into a tall narrow 
cylinder standing on a black surface. Several drops of dilute nitric 
acid are added (exact concentration immaterial), a stopper is appHed, 
and the vessel inverted several times in order to mix the solutions. ^ 
At this stage the Hquid should be clear when examined in a good side 
lighting. Five drops of dilute silver nitrate (17 grammes dissolved in 
a htre of disriUed water) are then added, mixed as before, and time 
allowed for any opalescence to appear. Under these conditions one 
part of chloride per miUion of solution may easily be detected in a 
column 10 cm. high. 

Drying metals 

Two methods are commonly employed for hastening the drying 
of metals — heating and desiccation. Heating in an oven at 105° C. is 
the simpler process, and is satisfactory for iron objects of all kinds. 
It is satisfactory also in the case of silver and copper alloys that are 
smooth and non-porous. Porous metal, however, takes longer to 
dry, and base silver and copper alloys that are at all porous may acquire 
a disfiguring film of oxide in the oven. Although oxide films may be 
removed by glass-brushing, this introduces another operation, and 

’ The use of a glass or rubber stopper is preferable to using the finger, as chlorides from 
the skin would be liable to upset the results. 



one that can be avoided if drying is done in a desiccator. Drying can 
be accelerated by passing an object through a bath of acetone and 
then placing it in a desiccator from which the air can be evacuated. 
A capacious vacuum desiccator is a very useful piece of apparatus to 
have in a conservation laboratory. It is charged with siUca gel (see 
Appendix XI), and evacuated as required by means of a vacuum 
pump attached to the main water supply (see Fig. 5, p. 152). The 
vacuum desiccator is used not only to dry objects, but also to store 
them until such time as they can be lacquered or protected with wax. 


When a metal object is showing signs of active corrosion, there 
can be no question that the only way of checking the activity and 
effecting a permanent cure is to employ chemicals, but this involves 
using certain ‘mechanical’ operations as well, in order to facUitate 
the action of the chemicals and their removal at the conclusion of the 
treatment. Mechanical methods may be employed directly for clean- 
ing oxidized or tarnished metals, where it is merely a question of 
removing surface staining; they may also be employed in dealing 
with rust fragments that are completely oxidized, and no longer 
subject to chemical change. 

Even with the most primitive workshop facihties it is possible to 
make a simple kit of tools that will be adequate for most purposes. 

Making needles and chisels 

The commonest tools are needles and small chisels, and the most 
satisfactory are those that one makes oneself. Needles may be mounted 
in wooden or metal handles, or in a small chuck or mandrel. They 
should be short and stiff and of varying thicknesses. An etcher’s 
needle, as sold for dry-point work, is a useful instrument for general 

A series of chisels and scrapers can be made from old files that have 
been annealed in the fire, or from pieces of silver-steel rod of round, 
square, or rectangular section. The cutting edges are made first of aU 



by grinding, and then the tools are given heat-treatment to adjust 
the hardness prior to sharpening. The heat-treatment is conducted 
in two stages as follows: 

Hardening. Heat about i inch length of the chisel at the sharp end 
to bright redness, avoiding over-heating, then immediately quench 
the metal by dipping the point in water and moving it about rapidly. 
This win make it dead-hard and brittle, and it wiU be dark in colour. 
Pohsh the metal on a fine emery paper until it is silvery white in 
order that the temper colours may be observed in the subsequent 

Tempering. To toughen the hardened metal, heat the pohshed 
chisel J inch or even farther from the cutting-edge — very gently 
and under strict control — and watch the temper colours moving 
towards the cutting-edge in the sequence — yellow, straw, purple. 
Aim to arrest operations by throwing the tool into cold water when 
a middle-straw colour extends from the cutting-edge some distance 
up the chisel. If the colour of the cutting-edge has gone to the purple 
stage, the chisel has been overheated, and the hardening and temper- 
ing operations must be done again. * 

Finally, clean and sharpen on a carborundum shp or oilstone. 

Mechanical methods may be defined as picking, chipping and 
scraping, grinding, cutting, brushing, shot-blasting, poHshing, and 
bumishing.2 These will now be considered seriatim: 


Picking is a technique apparently so commonplace as to seem un- 
worthy of further description, but there is much more to it than 
would at first appear. It is not often reahzed that a force of i lb. 
exerted by means of the point of a needle 1/5000 inch in diameter 
is equivalent to a pressure of several tons to the square inch. 

' A pale yellow colour represents razor hardness and the metal is brittle; a blue or 
purple colour impHes flexibihty as required for springs and hack-saws. Intermediate straw 
or brown colours are an indication that the steel will be tough enough for use as a chisel 
and wJl keep its cutting-edge. 

^ See Maryon, H., Metalwork and Enamelling, 3rd ed. (revised), 1954. 



A needle is thus a very potent tool and the correct nethod of handhng 
it is shown in the Vagram (Fig. 7). If pressed near the edge of a 
brittle surface, the edge will chip off, and the nearer to the edge it is 
pressed, the less the pressure required to remove tiny flakes from the 
rust escarpment. Considerable masses of rust can be eliminated, 
grain by grain, by using a needle in this manner; this is a labour of 

Fig. 7. Mechanical Removal of Rust: 

(a) Needle held at the correct angle and in the right place (at edge of oxide layer) 

(B) Needle held at the correct angle but in a position hable to fracture the metal core 
(c) Angle and position of needle both incorrect 

patience and the operator must be content to advance very slowly, 
using the minimum of pressure. Any attempt to hasten proceedings 
by pressing the needle too far from the edge, places unnecessary strain 
on the object and may end in disaster. The needle may be used on 
occasion as a tiny lever or even as a scraper. 

Chipping and scraping 

When metalhc excrescences have to be removed this can be done 
by using a small chisel and jeweller’s hammer, or a chisel mounted for 
use as an engraving tool. Mechanically operated needles and micro- 
chisels are available, but, owing to their vibration, these are only 
useful for heavier work. 

When the object is very thin and brittle, chisels are not employed 
as they are Hable to fracture the specimen: it is safer in such a case to 
use the needle or fine engraving tool, but dental burrs may at times 
be useful to help to reduce thick lumps of oxide. 

When incrustations have been softened by reduction, it is useful 
to be able to remove them in the wet condition, and in cases where 
it would be unsafe to use metal tools, lest they should cut into the 



surface, a sharpened stick may be employed, cut to a chisel edge. 
Cane tools are also useful as scrapers, and these may be used to rub 
the surface of an object undergoing treatment with zinc and caustic 
soda, in order to observe the progress of the reduction. 


When excrescences are exceptionally hard, it may be preferable to 
grind them away. For this purpose a small wheel or inverted cone 
may be used on a dental driUing-machine; alternatively, the grind- 
ing may be done with sHps of carborundum which can be obtained 
in various shapes and degrees of roughness. Grinding, however, must 
not be hghtly undertaken, as it can destroy markings in the rust — 
textile marks, wood grain marks. See., which may be of significance. 
On the rusted blade of a sword, for example, they may indicate that 
a scabbard had been present and decayed in situ; or markings on the 
tang of the weapon may be all that remains to show how the hUt 
was constructed. It is also necessary to add a warning that grinding 
may give iron oxide a dull metalhc poHsh so like that of metaUic 
iron itself that it becomes difficult to distinguish between the two. 


Cutting by using a hack-saw is only employed as a means of separ- 
ating two objects that are corroded together into a soHd mass, and 
which could not otherwise be separated. Such a mass may be cut as 
a preliminary to chipping and picking. Sht-saws are sometimes use- 
ful, or fi-et-saws, or even a copper wire hardened by stretching, fixed 
in a firame and fed with carborundum powder in oil. The quillon 
of a sword that had sHpped along the tang in antiquity and was 
rusted firmly into position in the wrong place was released by work- 
ing a small hole through the rust between quillon and tang, inserting 
the copper wire and ‘sawing’ a sHt through the rust along the tang. 
The wire was found to be much more effective for the purpose than 
the finest firet-saw since it was more flexible. 


Brushing is a frequent accompaniment of all other mechanical 



processes, its object being to eliminate foreign matter and reveal the 
progress of the work. Where a good sohd iron core remains and 
objects can be dealt with chemically, occasional brushing in a stream 
of water is essential to eliminate the by-products of the reaction, 
which tend to adhere to the specimen in the form of a sludge. 
Different kinds of brushes can be selected to suit the type of work in 
hand. Hard nylon tooth-brushes are invaluable for use on bronzes 
after reduction. If a steel brush is used, it must be dried thereafter or it 
will rust. A motor-driven brush can lead to a great saving of time 
when objects are strong enough to stand this treatment. When wire 
brushes of appropriate coarseness are used, either hand or motor- 
driven, they will help to scrape away loose rust, but care must be 
taken that the bristles do not penetrate too deeply into softer parts of 
the incrustation. Rotating wire brushes, if soft enough, ^ smear the 
products of reduction over the metal and make the surface appear- 
ance more uniform. In some cases it is safer to use graded steel wools 
rather than brushes. The finest grades are ver)^ gentle in action, es- 
pecially when used with a httle oil. Glass brushes are used for fine work 
such as jewellery or for polishing silver inlays in iron — they are in- 
dispensable in the laboratory. For removing hard rust from soft metal 
inlays, there is no more effective tool than a carborundum pencil, or a 
sUp of this abrasive that has been rubbed down to a point. It enables 
the rust to be pushed off without rubbing the surface of the inlay. 


One of the most effective methods of dry-cleaning is to use a jet 
of particles of fine grit blown from a modified form of spray gun, 
as in the shot-blast cabinet. Several types of cabinet are available 
commercially. Fine work may be cleaned in the dental cabinet, using 
a spray of bauxite grit, and at the other extreme, heavy metal castings 
are dealt with by using various grades of powdered cast iron, but 
bauxite is suitable for most museum purposes. It is possible by vary- 
ing the grade of bauxite, the pressure, the distance of the nozzle from 

’ A warning is necessary against the indiscriminate use of brass brushes as they tend to 
impart a yellow brass colour to other metals. 



the object, the angle of attack, dec., to carry out a variety of cleaning 
operations quickly and efficiently. This type of cleaning is never used 
where soft metals are present as it would give them a matt surface. 
It is useless, therefore, for gold, silver, and lead alloys, but it can be 
ver)' useful for the removal of mud and soft corrosion products ftom 
bronze and iron. Indeed it can safely be employed for the removal 
of sihceous deposits and dirt ftom certain kinds of smooth bronze 
patina without harming the surface. 

Shot-blasting has also proved to be the best means of removing 
surface incrustations of various kinds from baked clay tablets (p. 321). 

Polishing and burnishing 

Pohshing is done with the finer grades of abrasives such as emery- 
flour, rouge, diamantine. These are used either as powders or made 
up with taUow in paste form. Special cloths impregnated with such 
materials can also be obtained. Abrasives are available in emulsified 
form as metal pohshes, or compounded with paraffin and cotton- 
wool as pohshing wools. The degree of abrasion depends not only 
on the size of the particles, but on the nature of the abrasive — car- 
borundum powder, for example, being harder and having a greater 
cutting action than pumice powder of the same degree of fineness. 
Levigated alumina is a finishmg abrasive by means of which a high 
pohsh can be obtained on metal that has been previously rubbed up 
with a series of coarser grits. The final effect of brflhance can be 
obtained on steel only by burnishing to ehmmate all porosity from 
the surface. This is done either by using a cloth wheel on a buffing- 
machine or, in the case of fine work, a hard poUshhig-stone, e.g. 
haematite, the so-called bloodstone. A chain burnisher is very seldom 
used, and then only for coarse work in exceptional cases. Burmshing 
and pohshing by abrasives, although of great importance in cleaning 
jewellery, or in removing rust spots from otherwise bright armour 
and accoutrements, are, however, techniques of hmited appUcation. 
In cases where the rust is not of a spotty nature, but present as a con- 
tinuous layer of fine dust on a coat of mail, for example, it should 
be removed as far as possible with a stiff brush, if the metal is strong 



enough to stand this. No attempt should be made to attain a bright 
finish by burnishing with an abrasive, as the hnks will almost cer- 
tainly have been weakened by rusting. In such cases sufficient pro- 
tection is given, after cleaning, by the apphcation of a thin film of 
wax, which, when brushed, leaves a dull shine of satisfactory appear- 
ance that acts as a foil for any embellishments. This type of treatment 
and finish, by brushing, waxing, and poHshing, has been apphed with 
considerable effect in cleaning sword guards and t’suba where the 
steel is covered with fin e ornament and often inlaid with gold and 

The most perfect examples of the armoiurers’ craft in existence are, 
without question, Japanese swords, and if these become spotted vdth 
rust, the use of tools or of etching or abrasive materials in an effort 
to remove the rust does more harm than good. AH that can be done 
is to try to mininme the disfigurement caused by the dark spots of 
rust on the mirror-hke surface of the steel by alternate apphcations 
of kerosene and lanohn, and by rubbing with silk or a Selvyt cloth. 
The surface may be protected thereafter with a thin film of a neutral 
microcrystalline wax (see Appendix XII). Care is necessary to avoid 
possible accident, as the swords are often exceedingly sharp. 


The general methods of treatment that are apphed in dealing with 
corroded objects have been described, namely, reduction methods 
(electro-chemical and electrolytic), mechanical methods, washing, 
testing for chloride, and drying. The action of chemical solvents in 
arresting corrosion and cleaning and restoring metal objects is of 
equal importance as a general method of treatment, but, as solvents 
are selective in action, they are dealt with in the speciahzed chapters 
that follow. 



Gold is found in nature in the metallic condition as a rich yellow, 
soft metal, commonly associated with quartz and certain sands. It 
does not corrode, it is not dissolved under natural conditions, but it 
is often found alloyed with the baser metals silver and copper. The 
presence of silver in gold gives pale-coloured alloys in which the 
degree of paleness is in proportion to the amount of silver present, 
and such alloys have been known since the time of Phny as ‘electrum’. 
Both gold and electrum have been used for coinage and jewellery 
from the earhest times. 

When gold is found alloyed with copper, there is usually some 
silver present as well, and although such alloys may be very pale 
yellow or even greenish yellow when freshly cast or cleaned, they 
lose their pallor after years of burial in the groimd and acquire a 
warm yellow hue. This is known as surface enrichment and is due 
to the action of salts in removing the baser metals from the surface of 
the alloy, leaving a film of pure gold. 

It is usually desirable to preserve the rich colour, even though it 
may only be a surface enrichment, and in cleaning, care should be 
taken not to rub the surface because its quahty is easily lost. Incrusta- 
tions of hme can be removed from the gold by applying dilute nitric 
acid (i per cent.) locally with a match-stick, a small brush, or a capil- 
lary glass tube. SiHceous or muddy incrustations on gold may be 
removed by soaking in a 2 per cent, solution of a surface-active agent 
such as Lissapol. Organic remains are softened by immersion for a 
few minutes in caustic soda (2 per cent.) and removed with a match- 
stick. If incrustations resist these attentions, they may sometimes be 
disintegrated by using a fine jet of steam. 



Gold is sometimes found having a clear purplish red or rose-pink 
colour. In the eighteenth to twentieth dynasties of Egypt, a striking 
rose-pink colour was imparted to gold used for personal ornaments, 
horse-trappings, and sequins. On opening a box of royal robes in 
Tut-ankhamun’s tomb, for example, it was observed that the gar- 
ments had been decorated with sequins in a variety of shapes and, 
although the textiles had perished, it was clear that in one instance at 
least a garment had borne a stitched panel of massed sequins regu- 
larly positioned, rose-pink alternating with yellow gold, to give a 
diaper effect, thus proving that the pink colour was not an effect of 
age. A pink sequin from the Tut-ankhamun robe was analysed and 
found to contain in addition to silver and some copper, 0’85 per cent, 
of iron which was the colouring constituent. The pink colour was 
confined to the surface of the sequin back and front. 

Attempts to reproduce this coloured alloy were successfully carried 
out by R. W. Wood, I who fused gold with iron and obtained an alloy 
having the same colour and micro-structure as the sequins. If this 
were the technique appHed in the early dynasties of Egypt, it would 
be necessary to suppose that the iron was introduced in the form of 
a mineral from which it could be released when in the crucible, as the 
melting-point of iron is 1,535° C- An independent examination ofthe 
sequin problem made at the British Museum Laboratory has shown 
that when gold containing silver and copper is fused with iron 
pyrites and soda, some of the silver and copper combines with the 
sulphur in the pyrites and rises to the surface of the molten gold as a 
dross, leaving the gold alloyed with iron. When this gold is ham- 
mered into sheets and given heat treatment, it develops a superficial 
colour characteristic of the Egyptian sequins. Iron pyrites (M.P. 
1171° C.) has about the same melting-point as pure gold (M.P. 
1063° C.) and often has the yellow metalhc appearance of gold, for 
which it might easily be mistaken; indeed it has been called ‘fool s 
gold’. If by chance some of this material were inadvertently mixed 
with gold scrap for the crucible, it would be possible for the rose- 
pink colour to be developed accidentally. If such were the case, it 
' Wood, R. W.,y. E^ypt. Arch. 1934, 20, pp. 62-65. 

2y. ROMAN in A K I R OI HARK Itl Ul- (.1 ASS liLO W N I N'l'O A I' M( !■ O R A I' h 1) SI L V 1. R 1 R A M !■ WO R k 



would explain why this particular technique appeared sporadically 
in these dynasties and was later forgotten. 

Gold coins sometimes have what appears to be a variety of this 
same rich red patina, but here it is an acquired characteristic, the 
colour having developed by contamination and possibly heating 
during the hfe of the coins. At its best this adventitious red can be 
very handsome, and is regarded by collectors of coins as a valuable 
patina worth preserving, but it is easily rubbed off and when once 
lost it is virtually impossible to restore the rich colour by artificial 

Although gold is so durable it is a very soft metal and can be beaten 
out to form the thin leaves used for manuscript illumination and for 
applying to gesso. It can, moreover, be dissolved in mercury fairly 
easily to give gold amalgam. This substance has the consistency of 
butter and when rubbed over clean silver or copper, it adheres. The 
mercury may be readily volatilized by heating and the silver or 
copper thus becomes covered with a continuous layer of gold. The 
process is known as ‘mercury gilding’ or ‘fire gilding’. 

It was interesting to find both techniques used together on the 
dragon heads that ornamented the periphery of the shield from 
Sutton Hoo (a.d. 650) now in the British Museum. The original 
heads were executed in bronze that had been gilt by the mercury 
gilding process, but certain repairs had been made in antiquity and 
some of the heads had been replaced by models in gesso covered 
with gold leaf. The same type of repair was also evident on the large 
dragon shown in Pi. 23. 

Gold is likely to be found as decoration on armour and accoutre- 
ments, on bucUes, on bronze statuettes, especially in the eyes of the 
figures, and on silver dishes, in which case it may be Hmited to the 
embeUishment of reserves of ornament. Silver so decorated is said 
to be ‘parcel gilt’. This type of decoration was exploited in Persia 
during the Se^’uk period for the embeUishment of horse-trappings, 
rose-water sprinklers, incense-bumers, vases, dishes, and caskets. 
When gold is present in reserve ornament on silver, special care is 
required in cleaning because the gold is thin and the surface soft. 

B 6157 




and the gold reserves could be easily damaged by the cleaning 
materials used to remove tarnish from the surrounding silver. 


Metals that are gilt should not be reduced because of the danger of 
losing the gilding, but if the gold overhes sohd copper or bronze, it is 
sometimes safe to apply the Rochelle salt process (see p. 239) for 
cleaning the base metal. If the base metal is encrusted with corrosion 
products, however, and it is known that a film of gold exists beneath 
or among them, the only possible hne to take is that of mechanical 
cleaning, using needles under a binocular microscope at a magnifica- 
tion of some ten diameters. Very dilute nitric acid (i per cent.) can 
be used to clean the surface of the gold when it is exposed to 
view, but acid should not be used to soften the base incrustation, 
otherwise the ornament may become detached. To reveal the gold 
can be a long and arduous task, but the results are rewarding if the 
shape and decorative appearance can be recovered. 

When gilding is present on bronze statuary that is exposed to soot 
deposits and accumulations of dust, it requires occasional cleaning, 
and as dirt of this nature is usually acidic, it responds readily to wash- 
ing with dilute ammonia. The statuary should be well washed with 
water thereafter. This simple type of cleaning was apphed to the 
royal effigies in Westminster Abbey, and the results were very 


Objects of gold are often found in a distorted condition when they 
are dug fiom the ground, and there may be a strong temptation to 
try to fold the metal back into shape. This must be resisted as it is not 
the easy task that it appears to be. Pure gold is readily malleable, but 
when it is alloyed, as it usually is to some extent, the metal becomes 
brittle on ageing, and mechanical manipulation should only be 
undertaken after scientific examination has shown that the metal is of 
such a composition that reshaping can be considered a reasonably 
safe procedure in the hands of an expert. 


21 1 

Many examples of the restoration of crushed golden objects, by 
careful annealing and tooling back to shape, are provided by the 
material excavated at Ur of the Chaldees by Sir Leonard Woolley. 
One of the finest specimens is the wig-helmet of Mes-Kalam-Shar, 
made from a single plate of 15 carat gold, and weighing 2 lb. 4| oz. 
A more intricate reconstruction was that of the gold bull’s head 
forming the frontal ornament of a lyre (Pis. 24 A and b). This was 
brought to the Laboratory as a complex, held together by wax and 
bandages, so that the relative position of the various pieces could be 
studied at leisure. The gold mask, ears, and horns were first recovered, 
and as the gold, though badly crushed, was much thinner than that 
of the hehnet, it was correspondingly easier to manipulate. The 
original shape of the animal’s head was re-estabHshed, and having 
done this it was possible to reset the eyes and reassemble the various 
pieces of carved lapis lazuU forming the beard. 

The special tools used for such work are metal and wooden stakes, 
levers of various kinds, mallets of box-wood, hide and horn, and 
shaped sand-bags of leather. ^ 

’ Maryon, H., op. cit. 



Silver is a soft white lustrous metal which fuses at about 960° C. It 
is found locally in the metalhc condition (as native silver), but is 
more generally distributed in mineral form, two of the commonest 
minerals being argentite (silver sulphide), and cerargyrite (silver 
chloride), commonly known as ‘horn silver’. The sulphide is black in 
colour, and the chloride a dirty white or slate grey. The pure metal 
is malleable and ductile, and is capable of taking a high poHsh, 
and for this reason has always attracted the craftsman. It has 
enjoyed a deserved popularity in most countries of the world for 
objects of personal adornment and decorative metal-work. Gold 
and silver have been used for coinage throughout the ages, and for 
this reason the ratio of values has always been a significant figure, the 
relationship fluctuating with the scarcity of one or other metal. 
Partington I quotes a number of examples illustrating how the rela- 
tive values changed at different periods and in different locahties, 
observing that in the earhest periods of civilization gold was the 
commoner metal, because it could be obtained from rivers, whereas 
silver had to be obtained from the hills where it occurred, not as a 
surface deposit, but buried, often deeply, in the ground. 

Both gold and silver were in general use at Ur of the Chaldees 
in the Early Sumerian Period, at a time when silver was a rarity in 
Egypt. Indeed, silver was never available there in any quantity until 
the Graeco-Roman period, and it is not surprising, therefore, that 
the robbers who entered Tut-ankhamun’s tomb were found to have 
concentrated first upon the silver. 

' Partington, J. R,, Origins and Development of Applied Chemistry, 193 5 - 




On exposure to city atmospheres, silver is readily tarnished by the 
formation of a thin surface film of silver sulphide, and for this reason 
pohshed silver requires regular cleaning. The sulphur which tar- 
nishes the silver may be carried in. the air in the form of sulphuretted 
hydrogen from flue gases either industrial or domestic, its action 
being particularly noticeable in foggy weather. There are other, per- 
haps not so obvious, sources of sulphur contamination. It may be 
impossible to keep silver bright in places where there are vulcanized 
rubber floor-coverings or where a case has a rubber seal of the 
draught-excluder type. Cheap paints may contain volatile sulphur 
impurities that will tarnish silver, and even certain types of dry paint 
film may be potential sources of trouble. Casein is a common consti- 
tuent of water paints and this is subject to bacterial action that releases 
sulphur compounds that will tarnish silver. Emulsion paints should 
not be used to decorate cases that are designed for the exhibition of 
pohshed silver unless they can be guaranteed as fiee from casein. Paints 
based on hthopone are also suspect. The safest paints to employ are 
either cellulose paints or good-quahty oil-paints based on titanium 
white. These should contain enough lead-driers to act as a sulphur 
barrier and prevent any sulphur compounds that may be present in 
the oil film contaminating the air in the case. A more insidious source 
of trouble is contact of the silver with textiles that have been ‘finished’ 
by treatment with chemicals containing sulphur. Such textiles may 
contain as much as 200 parts per milhon of reducible sulphur, whereas 
all that is required to cause visible discoloration is 2 parts per milhon. 

Tarnish is easily removed with plate powder or rouge cloths, but 
for antique silver which requires regular cleaning such abrasives are 
not recommended as they would tend in time to cause damage by 
wearing down stamp marks and fine ornament. The same apphes to 
Sheffield plate or silver-plated ware, where drastic cleaning or the 
use of chemicals might result in exposing the underlying base metal. 
In these cases it is safer to confine routine cleaning to rubbing with a 
soft cloth and a httle French chalk, moistened to a paste, if desired, 
with methylated spirit containing a few drops of ammonia. 



For silver that has been neglected and is heavily tarnished, the 
following simple electro-chemical method of treatment may be 
used. Place the silver in contact with aluminium, cover it with a 
dilute solution of caustic soda or sodium carbonate (washing soda) — 
a solution of 5 per cent, strength is ample — and allow it to remain 
until the stain has disappeared. It can then be washed in running 
water and pohshed with a soft cloth or cotton-wool. Alternatively a 
chemical solvent for tarnish is available called Silver Dip,i and this 
gives excellent results provided the silver is not left: in the solution 
longer than necessar)" and is washed thoroughly afterwards. 

Tarnishing may be prevented by the exclusion of sulphur in any 
form. This is easier to achieve with silver in storage than when silver 
is on exhibition, as protection may be afforded by wrapping in 
several layers of soft tissue paper, and if the silver has been polished 
and washed beforehand it will be hkely to remain bright for a long 
time. Alternatively, the silver may be kept in Polythene bags having 
an air-tight seal. A recent advance has been the introduction of an 
anti-tamish paper — a soft velvety paper that has been impregnated 
with chemical substances (copper compounds, chlorophyll, &c.) so 
that it will absorb hydrogen sulphide from the air, thus leaving cor- 
respondingly less sulphur to attack the silver. Anti-tamish papers 
should not be in actual contact Avith the silver or there may be stain- 
ing by transfer of the sulphur, but as external wrapping papers they 
cannot be other than advantageous. 

Silver should be exhibited in a dust-free case, provided with a 
breathing channel to allow for the expansion and contraction of the 
air that occurs with change of temperature. This channel which is 
usually filled with cotton-wool should contain in addition a loose 
plug of anti-tamish paper. The efficiency of such papers when placed 
actually inside a vitrine has been found to vary, however, because 
they tend to become easily desiccated. Actually they are more effective 
when freely exposed to the varying temperature and humidity of the 
atmosphere; nevertheless, when a bulk supply of anti-tamish paper 

' Design and Research Centre for the Gold, Silver and Jewellery Industries, Bull. 
15, 1953- 



can be included in a closed case, it reduces the rate of tarnishing even 
under dry conditions. This has been estabhshed in a series of tests in 
which the protected silver was formd to remain bright for up to 
a period of several months after unprotected silver had visibly 

Where silver is displayed in a closed glass case on shelving draped 
with Shantung silk that has been passed through a 10 per cent, 
solution of lead acetate, ^ dried and ironed, it may remain bright for 
many years, as demonstrated by Bengt Bengtson in the Nordic 
Museum, Stockholm. 

With certain designs of vitrine it is possible to arrange a tubular 
filter-vent, through which sulphur-free air can be introduced by 
blowing from outside to replace the contaminated air in the case. 
The joints of the vitrine are then sealed with Polythene tape. A suit- 
able blower can be improvised from a vacuum cleaner, the bag being 
replaced by a filter tube charged with material that will absorb 
sulphuretted hydrogen, e.g. pumice granules that have been previously 
treated with copper or lead acetate and glycerol. 

It will often be more convenient to protect silver objects by a thin 
coating of impervious lacquer and if this is apphed with care it is 
practically invisible (p. 229). 


In silver objects there is generally a clear-cut distinction between 
surface deposits that are disfiguring and those that are thin and 
stable and which improve the appearance. Tarnish is a disfigure- 
ment, but one occasionally comes across specimens of ancient silver 
having a thin, uniform, shiny black patina of silver sulphide, and this 
is worth retaining for its decorative quahty. It has not been possible 
by laboratory methods to reproduce such a black surface, which is 
probably to some extent an effect of ageing. 

A tbin film of silver chloride can itself form the basis of an agree- 

’ This method was fiKt suggested by Professor J. Tandberg; the lead acetate 
solution should contain 2 per cent, of glycerol and about i per cent, of ammonia. 



able patina, as in the case of silver objects excavated from soils con- 
taining Httle chloride. T’ang silver from the loess has sometimes a 
beautiful patina of a chalky texture and pale lilac colour, stained 
pink and green with cuprous oxide and malachite respectively. As 
this patina is stable and enhances the appearance of the object, it is 
desirable to preserve it. If the patina is concealing fine ornament, a 
compromise is necessary, however, and it may then be permissible to 
expose the ornament by rubbing with a mild abrasive. The impreg- 
nated cloths that are sold for poHshing metals are suitable for this pur- 
pose (Dura-ght, &c.) provided the appropriate grade of cloth is used, 
i.e. one for cleaning silver: those sold for cleaning copper and brass 
are too drastic in action and would abrade the metal unnecessarily. 

When chloride is present but sulphide preponderates, the silver is 
dark in colour and has usually a stained and corroded appearance. It 
may sometimes be improved by applying selective solvents (p. 219) 
to remove the staining, and by picking or brushing as required to re- 
lease any incrustation. Such treatment will restore the natural white- 
ness of the metal, but care must be taken that ornamental details are 
not sacrificed in the removal of the incrustation. It is better to take no 
action than to risk damaging the object; specimens in this condition, 
although unsightly, are generally stable and treatment is only war- 
ranted if it improves the general appearance and reveals ornament 
more distinctly. Corrosion by sulphide formation may sometimes be 
found to have proceeded to the hmit, as in the case of a Persian silver 
anklet examined by the author. In such cases the metaUic quahty is 
entirely lost, the material is black and brittle and has the appearance 
of old leather. In this condition the silver is beyond restoration. 

W^hen silver has been buried for long periods in salty soil, as in the 
case of the silver vessels from the Ziggurat at Ur, it is largely trans- 
formed on the exposed surfaces into stable silver chloride (horn 
silver), which has a grey earthy appearance of no aesthetic value. 
Deformation may be considerable, owing to the great expansion 
accompanying the change from metal to mineral, but as it is stable, 
treatment is only required to improve the appearance of the specimen, 
to expose details of ornament, or to recover an elegant shape. Horn 



silver is tough, and the object is robust so long as the incrustation of 
chloride is in position, but if removed or reduced, it should be home 
in mind that the vessel may be much weakened. Reduction tends to 
give a discontinuous surface film , lacking in coherence and rigidity. 
Another factor of importance is that stratification is a feature of the 
incrustation resulting from long-continued corrosion, and it is likely 
that some of the segregated constituents are in the form of dis- 
continuous strata. This is clearly seen in the sectioned silver bead 
from Ur (Pi. 25), where, in addition, the residual uncorroded 
metal in the heart of the object is of varying thickness and porosity. 
Accordingly, when an object of high silver content is so heavily cor- 
roded that httle silver remains (as, for example, in the case of a thin 
metal jug or bowl), treatment, if applied at all, should be carried 
out with a view to exposing surface features and removing irregulari- 
ties. Electro-chemical reduction would be likely to be too drastic in 
its effects. On the other hand, there may be every justification for 
reducing a massive silver object, where it is clear that a sohd core of 
silver remains to ensure the mechanical strength of the specimen 
after treatment. The strength of such a silver core is, however, not 
always easily assessed — old silver is invariably much more brittle 
than silver that has been recently cast. The brittleness is caused by a 
change in the microcrystaUine structure -with the passage of time, a 
change that takes place irrespective of whether the metal is super- 
ficially corroded or has remained bright and free from chemical 


Brittle silver may be toughened either by heating to a dull red 
heat, or by maintaining it at a somewhat lower temperature for some 
time. The operation requires dexterity and experience, because silver 
melts at about 960° C. and there may be impurities present that 
lower the melting-point of the metal considerably. The metal begins 
to glow at about 600° C. If there are any mineral compounds of 
silver and copper present, such a temperature cannot be attained 
without fusion. Silver chloride, for example, melts at 455° C. and 



cuprous chloride (nantokite) at 422° C. Where douht exists, or 
where the silver is very thin, the minimum of heat should be used. 
Toughening actually begins to take place at about 250° C., and if the 
silver is placed in an oven, the temperature of which rises in about 
two hours to 400° C., this should suffice for most purposes. 

In the case of sound metal, free from salts, the silver may be heated 
in an electric fiumace, the temperatme of which is adjusted to 600- 
650° C. The best results are obtained in the electric furnace: if a gas 
furnace is used, it is difficult to prevent the silver from blackening 
unless all fumes can be excluded during heating. When silver is 
debased with copper, heating will cause the formation of a black 
surface film of copper oxide. Staining due to oxidation is, however, 
easily removed in a bath of sulphuric acid (5 per cent.). 


In the case of silver that has been crushed out of shape, heat treat- 
ment is always apphed before any attempt is made to bend the metal. 
When once toughened the metal may be worked back gradually to 
shape with the fingers, or with wooden tools covered with chamois, 
a process requiring much patience because the softened metal hardens 
again on cold working, and repeated heat treatment may be re- 
quired as the work proceeds. When dents have to be removed from 
modem silver, there is more latitude, and it is usually possible to rub 
out small deformities, without preheating, by the use of shaped tools 
manufactured for the piupose. The most useful are bloodstone 
burnishers mounted in brass sockets on stout wooden handles. There 
is no question of using these on old silver, however, until it has been 
previously toughened. 

The largest of the silver salvers from the Sutton Hoo Ship Burial 
was excavated in a crushed and very brittle condition. It was tough- 
ened by heating it repeatedly with a large blowpipe flame and 
restored to shape with the fingers, and later with the help of a hide 
mallet. When a free flame is used in this way heating should be car- 
ried out in a darkened room, as it is impossible otherwise to detect 



when the metal begins to glow, and part of the object might easily 
be inadvertently over-heated. 


Reference has akeady been made {p. 189) to the cathodic protection 
of silver in the presence of corroding copper. A common result of 
such action is the formation of a heavy deposit of copper corrosion 
products on the silver that is itself uncorroded, and this may be 
difficult to remove. The usual procedure is to get rid of as much as 
possible of the incrustation by mechanical methods (picking, grind- 
ing, &c.) first of all, and then to endeavour to eliminate the rest by 
the action of selective solvents. Some examples will now be given 
to illustrate the treatment of silver objects by the apphcation of 
solvents, selected because of the specific properties they possess. 

(i) Ammonia. In the presence of air ammonia dissolves copper and 
it may be claimed that no reagent is more effective for cleaning silver 
that has become encrusted with a massive deposit of copper by long 
contact with the baser metal. Ammonia, however, must be used with 
restraint where silver chloride is present, because it dissolves this sub- 
stance and, if apphed to corroded silver, would leave the metal in a 
much weakened condition. It softens organic remains such as residues 
of wood, skin, textiles, &c., and is thus a general cleaning-agent where 
such types of incrustation are concerned. It is used in concentrated 
form (density o-88), preferably in a firnie cupboard. 

(ii) Potassium cyanide. This is highly poisonous and seldom re- 
quired, but it is a very potent stain remover. It is not apphed in the 
form of a solution and it should on no account be used where acid is 
present or be allowed to come in contact with the skin. For removing 
stains from silver, a small pellet of cyanide is held with forceps and 
rubbed on the wet silver, and in dissolving a httle of the silver it will 
at the same time remove most stains. 

(iii) Formic acid. This has the particular merit that it may be used to 
dissolve copper compounds without affecting silver or silver chloride. 

(iv) Silver nitrate. For removing a thin film of metalhc copper 



from silver, the use of a solution of silver nitrate is preferable (see 
p. 224). 

In certain circumstances it may be necessary to use more than one 
solvent. Thus, for example, in the case of the silver lamp (Pis. 22 A 
and b) the bulk of the copper incrustation was cut away but a resis- 
tant layer of metalhc copper remained firmly adhering to part of the 
frail silver. The removal of this necessitated prolonged treatment in 
ammonia followed by local treatment with potassium cyanide. After 
all traces of the incrustation had been removed the lamp was dark 
in colour; it was given a bath in warm formic acid (about 30 per cent, 
strength) for a short time (twenty minutes) to brighten the metal 
and, after washing and drying, the silver was toughened in the 
furnace as described above. 

A striking illustration of the use of formic acid was in the treat- 
ment given to a silver cup from Enkomi, Cyprus (fifteenth cen- 
tury B.c.),^ which arrived at the Laboratory covered with a green 
incrustation. This incrustation had originated partly from the copper 
constituent of the silver alloy and partly no doubt from contact with 
base metal in the ground, and it gave the impression that the cup was 
made of bronze (Pi. 27 a ], conceahng the fact that it was actually of 
silver elaborately decorated with niello and gold. The true nature 
of the object having been discovered by X-ray examination, formic 
acid was selected as the cleaning agent, and after a successful experi- 
ment on one of the fragments, the cup was immersed for twenty 
minutes in a boiling solution composed of one volume of commer- 
cial formic acid and two volumes of water. By such treatment the 
green incrustation was removed, leaving patches of cuprite and re- 
duced copper which yielded eventually to local apphcations of 
ammonia (o-88). The ornament around the cup was revealed in 
surprising freshness, the niello for the most part being in excellent 
condition and the gold inlays rich in colour without showing any 
evidence of blanching (Pi. 27B). The silver proved on analysis to be 
alloyed with 3 -4 per cent, of gold and 9 per cent, of copper. 

Niello, a black decoration, has been used for ornamenting silver 

’ Schaeffer, C. F. A., Enkcmi-Alasia (Excavations 1946-50), 1952, Klincksieck, Paris. 



from very early times. Plinyi refers to the technique having been 
employed in Egypt, though few examples have survived. There are, 
however, black decorations still extant on the dagger-blades from 
Mycenae dating to 1500 b.c. and the niello technique is well estab- 
Hshed in Roman, Byzantine, and Anglo-Saxon art. Examples of 
black inlays from many sources have been subjected to laboratory 
examination by Moss^ who has developed micro-tests for their 
identification and shown that prior to the eleventh century niello 
corresponded with the mineral acanthite (silver sulphide), whereas 
subsequently a fusible mixture of silver and copper sulphides was 
used that corresponded to the mineral stromeyerite. Lead sulphide 
(galena) is sometimes present in addition. 

Niello ornament might be difficult to discern on heavily tarnished 
silver, or might even be echpsed by the spread of incrustation; it is 
necessary, therefore, to add a warning that the apphcation of any 
electro-chemical process would reduce the black surface of the niello 
to metal, and thus destroy the decoration. Should doubt exist, 
electro-chemical methods are best avoided; but if the reduction has 
been brief, the black colour of the niello can be exposed again by the 
judicious use of metal poHsh, the reduction effect being purely super- 
ficial owing to the high electrical resistance of the nieUo relative to 
that of the silver. 

(v) Ammonium thiosulphate and thiourea. A useful process for the 
removal of thick horn silver (silver chloride) is to glass-brush^ the 
surface with a 15 per cent, aqueous solution of ammonium thio- 
sulphate containing i per cent, of Lissapol. The thiosulphate may be 
replaced by 5 per cent, thiourea although this requires to be rubbed 
more vigorously and is best reserved for cast silver such as buckles 
and ornaments that are rigid enough to withstand the mechanical 
stresses involved. The thiourea method is, however, very mild in its 
chemical action and can be perfectly controlled. It can be used where 

' Natural History, Book xxxiii. 131. 

^ Moss, A. A., ‘Niello’, Studies in Conservation, 1953, i, pp. 49-62. 

* If the brush is provided with a metal holder, it should be washed frequently otherwise 
the holder will soon become corroded. 



niello decoration is retained in horn silver, and for this reason alone 
it forms a very useful addition to the methods available for conser- 
vation. The silver buckle (Pi. 28) responded well to treatment by the 
thiourea process. 

(vi) Citric acid. This is a convenient cleaning-agent for corroded 
copper, and hence for base silver covered with green corrosion com- 
pounds. It has the merits of being non-poisonous and odourless, but 
it must be used with a certain reserve as it is a non-selective stripping 
agent, and even in 5 per cent, aqueous solution may tend to over- 
clean. It is used as a pickle for separating objects such as base silver 
or copper coins that have become corroded together (see below). 


Electro-chemical methods may he apphed to remove corrosion 
from silver within the limits already described in this chapter (p. 217). 
Besides the aluminium and soda method referred to for removing 
tarnish (p. 214), a number of other methods are available that give 
satisfactory results in dealing with corroded silver. For example, it 
may be that in using 30 per cent, formic acid as a solvent the cleaning 
has not been entirely effective; in this event, zinc or aluminium 
powder may be added and the mixture heated for a time. An alterna- 
tive to the formic acid and zinc is to use zinc and caustic soda, which 
has a more pronounced softening effect on the corrosion. The horn 
silver becomes slimy after a time, and the object considerably 
weakened; brushing during reduction must be gentle and frequent, 
especially towards the latter stages. The reduced silver chloride in- 
crustation takes the form of a white spongy deposit which is easily 
removed, and the residue can be consoHdated by burnishing with a 
glass rod to restore the metal-hke quahty. After electro-chemical 
reduction the specimen may be hung in an electrolytic tank for ten 
minutes or so and boiled afterwards in several changes of distilled 
water. The alternative is to allow it to remain under ru nnin g hot 
water for several hoiurs, but in any event the washing is completed 
in several changes of distilled water until all chlorides are removed. 
The intensive method of washing (p. 197) can also be apphed to silver. 



After treatment, the silver is dried in an oven at 105° C. and it 
should be kept there for some hours, as the surface of reduced silver 
tends to be porous. It may sometimes be preferable to dry in a vacuum 
desiccator (see p. 200), as oven-drying tends to produce a film of 
tarnish. Where corrosion has been heavy and deep-seated it may be 
imwise to attempt to toughen the silver as the necessary high tem- 
peratures would tend to cause distortion of the metal. 


Silver may be reduced in the electrolytic tank in the usual manner 
using caustic soda as the electrolyte and iron anodes. 

A variation, particularly to be recommended in the case of this 
metal, is to substitute formic acid of 15 per cent, strength for the soda 
electrolyte, this being the concentration giving highest conductivity. 
This method was very useful in dealing with the following rather 
unusual specimen. The silver casing that was attached to a Greek vase 
of blue glass was badly corroded, and as the vase was of such a shape 
that the silver could not be removed for treatment it was necessary 
to select a method that would clean the silver without harming the 
glass. Choice fell upon electrolytic reduction in 15 per cent, formic 
acid, using anodes of stainless steel, and the result was considered to be 
entirely satisfactory (Pi. 29). A caustic soda electrolyte could not 
have been used here as it would have been hkely to attack the glass. 

Stainless steel anodes are not an essential in using formic acid — 
indeed prolonged electrolysis in formic acid leads to the steel being 
attacked and there may then be a deposition of iron on the silver. 
With iron anodes in formic acid this deposition would take place 
even more quickly. There are no such comphcations when using 
anodes of graphite or carbon with formic acid. Good results are 
obtained, for example, using carbon anodes with an e.m.f. of 12 volts 
and a current density of i amp. per sq. dm. 

During electrolysis of a base silver object the copper minerals in 
the incrustation are reduced to metalhc copper, and when using 
the formic acid electrolyte there is a tendency for a film of copper to 



be left on the object. This film is removed (after heating in an oven 
to 110° C. to eliminate formic acid) by immersing the specimen in 
20 per cent, silver nitrate made up in distilled water. By this means 
the film of copper is replaced by silver powder which is easily 
brushed off. 


Old silver coins that have been found in the ground provide 
examples of decay which will vary according to the quaHty of the 
alloy and the nature of the sod. 

Identification of the coins is a primary requirement, and treatment 
is carried out with a view to making the ornament distinct and the 
inscription, if any, readable. It is sometimes the case, however, that 
an inscription is more legible in the corroded state, and for this 
reason it should be studied before treatment. After such preliminary 
study, a hoard of coins is assorted into groups according to the differ- 
ent forms of treatment required. In some cases corrosion may have 
proceeded in such a manner that the surface of the coin is black and 
wax-like, and very httle metal remains. This is characteristic of a sul- 
phide incrustation, and such coins had best be treated only by washing 
and drying. When the surface is cracked or the coin is very porous it 
may be reinforced by impregnation with a nitrocellulose solution. 

More usually, a preponderance of metal remains, and the incrusta- 
tion is due to silver chloride; but silver coins are generally debased 
and the incrustation will be found to be mixed with the products of 
the corrosion of the copper present in the silver alloy. The coins may 
then be adhering firmly together in masses. A conglomerate of this 
nature may be broken dovm by formic acid treatment or by soaking 
in 5 per cent, citric acid, although sometimes the individual coins are 
more easily released by reducing the mass with zinc and caustic soda. 
The alkahne Rochelle salt process (p. 239) has advantages where the 
silver is greatly debased with copper, but in this case tlie coins must 
be carefully brushed from time to time since the cuprite which 
cements them together is dissolved by the Rochelle salt exceedingly 
slowly and then only in the presence of air. 



In the BritishMuseum Research Laboratory, where coins frequently 
arrive for treatment in hoards, they are dealt with in a machine con- 
structed on the principle of the ball mill. This apparatus takes the 
form of a deep rectangular tank made of polyvinyl chloride mounted 
at an angle of about 45° from the vertical (‘tiunbler tank’) and 
arranged to revolve slowly about its longer axis. It is driven by an 
electric motor. For cleaning base silver coins the tank is filled with 
an appropriate reagent such as alkaline Rochelle salt, and a few loose 
coins are added to act as hammers when the tank revolves and so 
help to separate other coins from the conglomerate mass. As the tank 
revolves the contents fall four times from the flat walls during every 
revolution and these movements assist the chemical action so that the 
coins are soon separated. Individual coins are washed with water 
thereafter and dried in hot sawdust in the tumbler tank; they emerge 
clean and with a sUght polish which seems to heighten the contrast 
of hght and shade and thus enable lettering and ornament to be more 
easily studied. 

Before the installation of the tumbler tank, electrolytic reduction 
was employed, and while this procedure was found to give equally 
satisfactory results, it involved more labour. It was apphed, for ex- 
ample, in cleaning a hoard of 15,000 silver coins from Dorchester. 
During electrolysis the coins required continuous attention and had 
to be frequently removed and brushed in order to prevent the re- 
duced incrustations from being deposited as a resistant metaUic film 
on the coins. The small mass of corroded silver coins shown in 
Pis. 30 A and B had become agglomerated as a result of fire. The mass 
was loosened to a considerable extent by soaking for some hours in a 
bath of citric add (5 per cent.), final cleaning being done by placing 
the coins in stirrups of copper wire suspended from the cathode bar 
of the electrolytic tank. 


silver has been used so widely in decoration that all sorts of prob- 
lems are presented in the conservation of objects of a composite 
nature containing the metal in conjunction with other materials. 


B 6167 



In dealing with silver inlays it is safest to adopt the mechanical 
methods which have already been described (p. 200). If chemical 
treatment is essential, it may be necessary to isolate the inlay with a 
wall of Plasticine or clay, but there is always the danger of chemicals 
seeping beneath the inlay and presenting a problem in washing. When 
washing after treatment is impossible, as with a fretted silver orna- 
ment set in lacquer, cleaning can only be satisfactorily accomphshed 
by the use of abrasive cloths. The problem of revealing silver inlays 
in rusted iron generally resolves itself into adopting methods for 
removing the rust which tends to overrun and obscure the silver. 


An unusual and very interesting problem involving the conserva- 
tion of silver in juxtaposition to iron was provided by the Emesa 
helmet, a treasure fi-om Syria (Pi. 31), and as it presented certain 
novel features its restoration will now be described in detail. When 
received, this fine helmet, with its silver visor in the form of a human 
face, was at first taken to be a ceremonial piece, but, as pointed out 
by Seyrig,^ the protection afforded by the iron skuU-cap was con- 
tinued by a lining of iron of substantial thickness 1-6 mm. underlying 
the visor and conforming exactly to its shape. The probabihty is 
therefore that the hehnet was intended as head armour to protect the 
wearer under war conditions; the iron cranium was dented as if by 
a heavy blow. The silver was found on examination to be very brittle 
and cracked in many places; the major cracks had been filled with a 
dark stopping material in an effort to reinforce the silver as shown in 
the photograph. The iron backing of the visor had rusted and the 
swollen metal pressing behind the brittle silver had been responsible 
for distorting the features and forcing the cracks open, in some cases 
to as much as 4 mm. A state of strain had set in which, if not released, 
could only lead to further deterioration as the silver itself was only 
0-5 mm. in thickness. A necessary prehminary to treatment was the 
removal of the visor from the helmet; this proved to be easy because 

’ Seyrig, H., ‘Le Casque cl’£mese’, Annales, Arch, de Syrie, 1952, 2, Nos. i and 2. 



the hinge, although no longer movable, was already detached from 
the head-piece and the two parts were merely held together by wire. 
After temporary consoHdation by applying transparent SeUotape 
across the cracks in the mask, the visor was bedded in a thick pad of 
cotton- wool, and the layer of rusty iron was cut away from the back 
in the region of the mouth and lower jaw. It was in this region that 
the silver was weakest and most distorted through contact with the 
corroding iron. It was clear that no satisfactory restoration would be 
possible unless the silver could be toughened. A pilot experiment 
was carried out on a tiny detached fragment and as this gave re- 
assuring results the visor was heated in the electric furnace for three 
hours, the temperatme rising slowly during this time to a maximum 
of 3 10° C. Heating was arrested at this point. The silver was now 
darkened, pardy by the burning of the stopping of rust and wax that 
had been used in a previous restoration to fill the cracks and partly by 
the blackening of residual rust that remained attached to it. The rust 
was removed by brushing with 9 per cent, oxahc acid (see p. 278) 
before the final heat-treatment as additional heating would have made 
the iron compounds too insoluble to remove later. Further heating for 
eighteen hours at 600° C. and thirteen at 650° C. was necessary before 
the silver was considered to be sufficiently toughened to stand up to 
the manipulation that would be necessary in closing the cracks. In 
order to strengthen the visor, where iron had been cut away, the 
underside of the silver was cleaned and given temporary reinforce- 
ment by applying patches of silver gauze across the cracks with soft 
solder. The gauze took the shape of the silver very readily and could 
be fixed without involving any pressure. The front of the visor was 
then carefully cleaned with Dura-ght cloths to remove all of the 
dark oxidation products and it was then in a condition for the restora- 
tion work to begin. 

The first operation was to restore the shape of the lower jaw by 
dealing with the cracks, one series at a time. The temporary patches 
of silver gauze were removed, and the silver was manipulated to get 
the edges of the cracks to meet exactly, then fresh patches of gauze 
were appHed, and fixed permanently in position across the cracks on 



the underside, using soft solder with a non-corrosive flux. This was 
a long and exacting process but well worth all the care expended 
upon it because, as the work advanced and the cracks disappeared 
one by one, the visor became noticeably stronger and more rigid, 
and the likeness nearer to that shown in the earhest photographs. 

It should be emphasized that it was the toughening of the silver 
that made this type of restoration possible; if for any reason it had 
not been possible to toughen the silver by heat treatment, it could 
not have been manipulated. In its brittle condition the only possible 
action would have been to try to reinforce the inside, and fill the 
cracks ftom the front. It is fortunate that the more permanent type 
of restoration could be carried out on the Emesa visor, and that it was 
not necessary to use filling material, as this tends in time to become a 
conspicuous disfigurement. Care was taken to leave no solder on the 
front of the mask; it appeared inevitably as thin lines in the repaired 
cracks, and these were concealed by the local apphcation of a surface 
coating of stiver, appHed by ‘ragging’ wtith a pointed silver anode 
tipped with cotton-wool moistened with potassium cyanide (p. 289). 

It now only remained to replace the rusty iron that had been re- 
moved from the back of the visor. This material, after cleaning, was still 
in the form of sizeable fragments that could be returned to position in 
contact with the silver, where they were consohdated with Durofix. 

The major fractures that remain in the silver after restoration no 
longer weaken the structure, nor do they cause serious disfigure- 
ment. It would have been possible to deal with all the cracks, but 
this would have meant removing practically aU of the iron from the 
back of the visor, and, as much of the iron was in sound condition, 
this large operation was not considered justified. The interest and value 
of the object depend as much on the variety and perfection of the 
craftsmanship displayed in its construction as on its actual appearance. 
With the removal of the loose and swollen rust, where the strain was 
greatest, and the consohdation of the iron and silver in the lower part 
of the face, the helmet is now considered to be in a stable condition. 

The small holes beneath the eyes in the visor had been designed so 
that the wearer could look down to the ground in front of him, but 



evidently they were not enough, and it is amusing to find that they 
had been enlarged, in each case, by a narrow notch cut in the central 
lobe, no doubt in an emergency as this adjustment is not the work of 
a craftsman. 


Lacquers that are of the right quality and consistency are almost in- 
visible on silver if they are skilfully appHed. Many types are avail- 
able, and some of the best are those based either on polyvinyl acetate 
or the polymethacrylates. A lacquer that has been found in practice 
to have excellent protective properties, and that is almost invisible 
when sprayed, or carefully brushed on silver, is the commercial pro- 
duct known as Ercalene. Frigdene is also very good.' 

Before lacquering, the silver must be cleaned and degreased, and 
while cleaning can be done by any of the methods recommended, a 
word of warning is necessary as regards the washing process to 
remove grease. Many proprietary detergent preparations are avail- 
able today that are excellent for household cleaning, but some are 
entirely unsuitable for silver, as they contain phosphates which react 
with the silver to form a resistant brown stain. It is better to degrease 
the pohshed silver by washing it in hot soapy water, or the pure 
detergent, Lissapol, may be used; the silver is then rinsed in running 
water and dried. After degreasing, care should be taken not to spoil 
the surface by touching it with the fingers. When quite dry the silver 
is ready to be lacquered. 

Lacquer may be appHed by spraying, dipping, or brushing. The 
brushing technique is probably the simplest. A soft clean brush is 
used and the lacquer should be suflSciently thin to spread without 
leaving brush marks. It must not be allowed to collect as a thick film 
in any hollows where it would catch the hght and reveal its presence 
by a shine foreign to that of the silver. It is useful to have a solvent 
at hand which can be used, if need be, to remove any surplus. 
Lacquer can be prevented from collecting in hollows by stipphng 
with a brush or by using a gentle puff of air from a blow-ball. 

’ Both preparations are marketed by Messrs. W. Canning & Co. Ltd., London. 



S ^ 








u o 

4^ oft 

o ■§•« 
c o \ 

2 «■ 

2 2 ii 
- S-lS I 

•-•I-. ® 

y t ) fti Vj 

i-^-s n 

_U'Zi U T5 

o " 

•S > 

M ^ 


bo 55 

'§ § 
.2 3 
a a 

-Si ■ 

o 5 

j : 8 ■ 

rt .2 


bC « 

C P » 

; s 8- 

- ist '- 

S O o 

« rt * 
O M 
2 " 

>>-0 tH 

— o £3 A 

t 2 » 



•S u *** 

•3 0 ^ 



^ a.'H 0 

Cu S ^ 


"O 0 § CO 

1 ^ 




2 *« 
i lil 

2 s « 
0-6 ^ 

2 i 3 s 

^ Si ^ 

« — 

: 2 2 ^ 

- — tj •*« ^— 




In’ “-a 
- o.S p 


1 ""St 

0 2 0 

1 ij 0 



1 is uj 
(2 ° 0 

. ns 

0 i> 

.H >^S 



■~t na t< 

2 0^2 


S' o"^ 

— y (j*.^ 


0 a 2 

r3 0 

i i 'O 

n 3 

rt . 

S "w ^ 
” -i "5 ^*' 

s 2 s u « 
M O ^ 00 

2-“ I.S ^ 
gs - s -^ 

O (3 

. o . 


n ti 

O ,2 X M > 
H «> O 5 

O 2^' S ;2 
U Wi 2 





rs c 

- o ^- 

C /5 ^ 

-O ^ 

.N ^ V* 

“12 '3 -o ^ 
a G ^ 00 
^ ►r « 

O 2 2> a ? 

[. No treatment. 

!. (a) Warm soapy water; {b) Lissapol froth. 

I. Mild abrasives: {a) French chalk; (i) Rouge; (c) ‘Plate powder’; (d) Pumice or emery powder; (e) ‘Metal polish’; (J) Dura-glit; (g) Special cloths. 
.. Solvents: 




> S 





.U Q U 

.S r! 

o *• 

u — ' ' 


o ^ 

II ■§ 

8j 3 t! 

5 fc S 


« OiJS <s 

Va *2 cL 

<3 gf o 

> t) 

'ir.a q5s 

-’-N o O * *5 

• u *3 V) — 

M 3 ; 

^ 0 g j. "y 3 ^ 

•3 gS H.S 

o h o "llS -S 
IS a.-2.a3 "0 

c J S "2 -c 

3 “••O 2 2 0 60 

'3 O'h '1 a " .3 

fl 3 o t- 
3 o*c « c 
c IS •- ra O .S j< 






T3 n3 S ■ 
I -a 0.3 

- >• 

U rH ^ 

S.3 " 


rt 'j-i 


o 5 .'5 o 


<3 ""G* 



00 00 

S 3 




rt .O 

c c 

'S p 

.3 0> « 

< w w 





■3 S 
O. 2 

n M 


O 3 






^ o 

^ t> 


■3^ s 

a X 
w g 

til ft 

<■> .s 


3 _ _ . 

2 2 Q w 

t) ^ • 

S-3:2a, o£2.§ 

rt O 
u V> - 

*3 y.a 

a o ^ ^ 

.2 T3 *3 'P 
'3 rt 
C y U U 

-2 S S.S 

«S * 










=3 -Ti >' 
y ^.S 

2 « E 

■ts ^ 
'*7 o 

« pH 



o> .3 

tifi CLi 

f O 

s is 

bo 3 

f % 

•a 3 

fco J3 

o o 
o* 2 

«-> S 


2 _S 

•S Oa 

O rt 
Xli. v> 

tr p 


*^3 avH 





y p-^ 

< 04 



t: M 


•a 3 

3 ^ 


M ^ 





Copper derives its name from Cyprus which was the Roman source 
of the metal, aes cyprium being the ‘bronze from Cyprus’. It is, how- 
ever, an element, and it occurs in nature in the metalhc condition as 
well as in the form of many minerals, chief of which are cuprite 
(cuprous oxide), chalcocite (cuprous sulphide), chalcopyrite (copper 
pyrites), and the basic carbonates, malachite and azurite or chessyhte. 

Metalhc copper resembles silver in being sensitive to sulphur which 
causes it to become covered with a film of tarnish consisting of copper 
sulphide. It differs from silver, however, in being sensitive also to 
oxygen; pure copper oxidizes very readily when exposed to moist 
air, whereas, under similar conditions in air that is free from sulphur 
compounds, stiver remains unstained. Thus it happens that while 
silver objects in the museum tend to tarnish, copper objects are more 
usually covered with a thin film of oxide which confers on them a 
dull appearance. The oxide film does not increase appreciably in 
thickness with time and for this reason it may be regarded as pro- 
tecting the underlying metal; oxidation may be so sHght that the 
metaUic appearance of the metal is maintained and there is no dis- 
figurement; but if the metal is alloyed with tin, lead, or zinc and the 
constituents imperfectly mixed, the oxidized surface may be patchy 
and give rise to a rather unpleasant appearance. The restoration of 
the metalhc condition is not, however, a major problem as the 
thinly oxidized metal may be cleaned with metal pohsh or by im- 
mersing the object for some hours in dilute sulphuric acid (5-10 per 
cent.), washing afterwards, in each case, drying, and rubbing up 
with a soft cloth. But the brihiancy is soon lost and to preserve a 
burnished surface, even under museum conditions, it is usually neces- 
sary to cover it with an impervious film of lacquer. 



When buried in damp sod, copper soon loses its metallic appear- 
ance. The oxide layer increases in thickness, and cuprous oxide be- 
comes compacted into the purplish-red mineral known as cuprite; 
this, in turn, may become encrusted with basic carbonates that are 
green or sometimes blue in colour and that correspond to the minerals 
malachite and azurite. Such incrustations are stable when free from 
chloride and they protect the underlying metal from further cor- 

It does not follow, however, that all green incrustations are stable. 
Appearance has httle to do with stabdity, even though it be true that 
a thin coherent deposit is more likely to be stable than a heavy porous 
incrustation that woidd tend to behave hke a sponge and absorb 
moisture and soluble salts from the sod. When more than one metal 
is present, as in the tin bronzes, ^ it is only to be expected that incrus- 
tations will be more compHcated both in composition and structure 
and therefore more hable to contain and retain salts. Contamination 
is the nde rather than the exception because soluble salts are widely 
distributed in nature, in the sod as well as in the sea, and for this 
reason they are usually present in excavated material. 

It was noted in the case of sdver (p. 216) that when this is buried 
in salty ground it tends to become covered with an insoluble shed of 
stable sdver chloride. In the case of copper and its alloys, however, 
the presence of chlorides in the incrustation presents an acute problem 
from the point of view of conservation because an unstable cuprous 
chloride is formed; this cuprous chloride continues to react and there 
is progressive corrosion, even under museum conditions, with the 
result that the surface becomes powdery and spotty. So common 
are these features in corroding bronzes that the appearance of 
characteristic pale green powdery spots is referred to as ‘bronze 
disease’ whether occurring upon copper or on any of its alloys. 

The spots grow as a result of the nantokite (cuprous chloride) 
being transformed into atacamite (basic cupric chloride) by the action 

' Gettens, R. J.,J. Chem. Education, 1951, 28, pp. 67-71. In this paper the reactions that 
take place during corrosion of a tin bronze are described and the morphology of the 
incrustation illustrated in a series of photomicrographs. 



of oxygen. This change is facihtated hy the presence of moisture and 
it is not surprising that ‘damp’ has been mistakenly credited with 
being the cause of corrosion since serious outbreaks of active corro- 
sion have been observed to take place in damp weather. But as the 
features of hronze disease cannot be reproduced even tmder damp 
conditions when chlorides are absent, it is clear that moisture is only 
a subsidiary factor in the corrosion of copper and its alloys. The main 
objective, therefore, in conservation is to get rid of chlorides. Here 
we are faced with two difficulties. Cuprous chloride cannot be re- 
moved by mere washing with water. Not only is it insoluble but it 
is inaccessible as it occurs in greatest concentration, not in the surface 
spots, but in the deeper layers of incrustation. In actual practice, 
chemical methods can be applied to convert insoluble chloride com- 
plexes into soluble salts that can be removed by washing, but it is 
the inaccessibihty of the nantokite that introduces the serious prob- 
lem. A corroded bronze has a banded structure (Pi. 26), the metal 
core being surrounded by layers of massive cuprous oxide inter- 
spersed with powdery tin oxide, and these layers underlie an external 
shell of basic carbonates. Chlorides may be present in the green sur- 
face shell, but in a heavdy corroded bronze concentrations of cuprous 
chloride are found in the absorbent layers of tin oxide, and even in 
the micro-cracks and fissures that can be shown by metallographic 
techmque to exist in the interface between metal and cuprous oxide. 

In view of these facts it would seem to be outside the bounds of 
possibihty to remove all chlorides from a mineralized bronze with- 
out at the same time decomposing the incrustation, and, indeed, this 
is the only certain way to stabilize a bronze that is heavily corroded. 
There are many laboratory methods by which this may be accom- 
phshed, and they are generally referred to as stripping methods. But 
there are times when it would be quite unthinkable to sacrifice the 
patina of a bronze and when, if corrosion has set in, ways and means 
must be found to deal with the attack while preserving, as far as 
possible, the appearance of the specimen. 

In the study of ancient metals it wiU be noted that bronzes have 
survived for thousands of years even though their porous incrusta- 



tions have been impregnated with salts. The explanation must be 
that an equihbrium had been estabhshed between the corroded 
bron2e and its environment in the ground, and it is to this that we 
must attribute survival. Excavation destroys the equihbrium, how- 
ever, and exposure to new and very different enviromnental condi- 
tions provokes changes that are often profound. It may sometimes 
happen that a bronze can undergo a second acclimatization to the 
museum environment, without visual change, but it is more usual 
for further corrosion to take place in the process, and if neglected, 
equihbrium will in many cases only be re-established at the expense 
of complete disintegration. AH metal objects should, therefore, be 
sent for laboratory treatment as soon as possible after excavation. 

When electro-chemical methods can be adopted, the objects will 
be restored to a stable condition and to something resembling their 
original uncorroded appearance. But it will not be possible to apply 
electro-chemical methods in every case. Reduction methods are only 
apphed when a substantial metaUic core remains and when its mechan- 
ical strength is beyond doubt. In the absence of a metaUic core the 
aim win be to arrest decay by the use of selective solvents. When 
electro-chemical or electrolytic methods are out of the question, as 
in the case of a finely patinated bronze, the prospect of arresting 
decay without destroying the patina will depend upon the nature 
and distribution of the corrosion, and also upon what success can be 
achieved in locaHzing the action of the chemical reagents during 
treatment. When conditions are dry, chloride activity will be at a 
minimum and if bronze disease should break out and can be dealt 
with in its early stages, there will be the less chance of its getting out 
of hand. 

In this type of problem the following points are important — (i) 
to treat the spots in the early stages, (2) to dry the bronze well 
thereafter, and keep it dry, and (3) to be ahve to the necessity of 
giving further treatment in good time should further spots make 
their appearance. Methods of treatment for bronzes where patina 
has to be preserved are given on p. 242, but it will be reahzed that 
these methods do not ehminate all chlorides from within and beneath 



the incrustation. Such treatment must therefore be empirical to a 
certain degree, and there can be no absolute assurance that preserva- 
tion wiU be permanent, although many cases are on record where the 
treatment has proved to be successful. By these processes, chlorides 
are removed from the surface where the shell of iucrustation is most 
porous and weakest, leaving the more soHd crust of corrosion to seal 
up the remainder. It is a common error to beheve that a short cut is 
possible — that wax or other film -forming substance may be used to 
seal in the decay and so act as a protective layer. Experiments are 
described below which show this to be a fallacy. No impregnating 
medium has been found to be of any permanent value in arresting 
the progress of ‘bronze disease’, unless the spots are excavated in the 
first place or treated by chemical means to enable the chlorides to 
be extracted. 

Before proceeding to describe and recommend processes of treat- 
ment, it may be useful to refer to some observations made during the 
examination of a mineralized ‘bronze’ spear-head of rectangular 
cross-section from Ur of the Chaldees that could be sacrificed for 
experimental purposes. Although covered with a heavy green in- 
crustation, the object was apparently stable under museum condi- 
tions and there was no evidence that corrosion had taken place in 
recent times. Chemical analysis showed the metal core to consist of 
97-2 per cent, copper with less than i per cent. tin. 

The spear-head was sawn into a number of fragments roughly 
cubical in shape. The cross-section showed the structure to consist of 
a metalhc core enveloped in massive purple-red cuprite, on the sur- 
face of which had grown a green patina or incrustation of about 
3 mm. in thickness, mainly malachite (Fig. 8). The test pieces were 
dealt with variously as follows: 

I. Controls. Some cubes were kept as controls, and these broke into 
active corrosion within forty-eight hours of being cut. On examining 
the section, pale green powdery material was observed growing 
thickly along the mner margin of the outer green patina. This was a 
zone of nantokite (cuprous chloride) which had been sealed within 
the incrustation, but on exposure to air it was converted to atacamite 



(basic cupric chloride) and exhibited the typical appearance of bronze 
disease within this very short period of time. 

2. Washing. Other pieces were washed in distilled water with the 
greatest care until apparently free from chloride and dried by passing 
through alcohol and ether, and maintaining at 105° C. in an oven for 
several days. When these were exposed in a hnmid atmosphere bright 
green striae formed on the exposed edges of the outer green layer 

Fig. 8. Diagram showing section of rectangular 
spear-head of copper in an advanced state of 

foUowdng certain lines of the strata, and these could be shown by 
micro-testing with silver nitrate to contain chloride. There was no 
indication of any chloride in the central metal core or in the cuprite 
immediately overlying it, nor could chloride be detected on a longi- 
tudinal external face of the specimen.^ The chloride was, therefore, 
located within the green rind of incrustation but not on the outside 
surface, and it could not be removed, even where exposed on the 
cross-section, by the most careful washing alone. 

3. Reduction. A third series of test pieces was subjected to electro- 
chemical reduction with zinc and caustic soda or to electrolytic re- 
duction, washed free from chloride, and dried. Corrosion could no 
longer be induced to take place even when the samples were kept 
in a damp atmosphere. Reduction is therefore a cure for corrosion 
when apphed for long enough in conjunction with washing to 

’ Although there was no evidence of it here, it is quite common to find the attacking 
layer of cuprous chloride between the cuprite and the metal core. 



eliminate all chlorides. But this process involves the loss of the green 

4. Acid treatment. Specimens were treated by soaking them in citric 
acid (5 per cent.) and sulphuric acid (15-20 per cent.). The latter was 
more effective. It seemed to go a long way towards removing the 
chlorides in the course of a few days. There are practical difficulties 
in using it, however (see p. 241), and it is not recommended as a 
standard process. These acids act as stripping reagents and the green 
surface is lost as in the case of No. 3 above. 

5. Alkali bath. A test was made by immersing some pieces in sodium 
sesquicarbonate solution (5 per cent.). This did not remove the green 
incrustation or patina, but gradually extracted the chloride in the 
form of salts that could be dispersed and eventually eliminated by 
frequently renewing the solution. This operation, including washing 
with distilled water, took several weeks to reach the stage where 
chlorides could no longer be detected and the dried bronze was stable 
when tested under moist conditions. There can be no doubt that, after 
such treatment, much chloride must remain inaccessible beneath the 
green surface, but it seems that the sodium sesquicarbonate gives rise 
to basic copper carbonate which fills the pores of the incrustation, 
sealing off the residual chlorides from atmospheric moisture. 

Assuming that the test-pieces are representative of the average con- 
ditions that obtain when heavily corroded bronzes are contaminated 
wdth chloride, the experiments indicate in which part of the incrusta- 
tion chlorides are to be found in the greatest concentration, and how 
very sensitive they are to moisture. They prove the inadequacy of 
washing alone, the effectiveness of stripping methods, and the possi- 
bdity of protection by removing exposed chloride, and sealing the 
remainder behind a moisture-barrier. 

The choice of a method of conservation depends upon the physical 
and chemical condition of the specimen, the amount of chloride 
present, and the nature of the incrustation — ^its thickness, stratifica- 
tion, and porosity. If it is essential that the patina should be saved, it 
is clear that the operation may be a lengthy one, and the results 
inevitably fraught with some imcertainty. 



I. When incrustation or patina can be sacrificed 
When the incrustation or patina can be sacrificed, corroded copper 
alloys may be dealt with by electro-chemical (p. 191) or electrolytic 
reduction (p. 194), or by employing alkaline Rochelle salt. The latter 
process has been found to be the one most generally useful. Prehmin- 
ary treatment by Rochelle salt is recommended, therefore, for deal- 
ing with copper incrustations, even where apparatus is available for 
electrolytic reduction, though it will be found convenient to employ 
the electric current in the final stages, in order to fadhtate the removal 
of salts and chemicals, and thus economize in the time required for 
the final washing in distilled water. 

(i) Use oj alkaline Rochelle salt and dilute sulphuric acid. Copper com- 
bines with oxygen in two difterent proportions by weight to form 
two definite oxides, cuprous oxide and cupric oxide, and these com- 
pounds are commonly present in incrustations, and give rise to two 
series of salts, the cuprous salts and the cupric salts respectively. These 
may be carbonates, chlorides, sulphates, &c., and the chemical namre 
and distribution of these compounds •will depend on the conditions to 
which the metal has been exposed. Usually the cupric salts predomin- 
ate in the visible incrustation, and they happen to be soluble in an 
alkahne solution of Rochelle salt. By dissol'ving them away the green 
encrusted surface of a copper alloy is broken down, and it then 
remains to find a method of eliminating the residual cuprous salts. 
This can be done with dilute sulphuric acid. There are two operations, 
therefore, and these are apphed consecutively in the standard form of 
Rochelle salt treatment, the result of their appHcation being to re- 
move all the minerahzation and to restore the metaUic condition of 
the specimen. 

The tw^o solutions are prepared as follows: 

Solution 1. Alkaline Rochelle salt. Dissolve i oz. of commercial flake 
caustic soda in a pint of cold water, and dissolve 3 oz. of Rochelle 
salt (sodium potassium tartrate) in this solution. 

Solution 2. Dilute sulphuric acid (lo per cent.). Add 2 oz. of 



concentrated sulphuric acid slowly to a pint of cold water in a por- 
celain dish, stirring continuously d 

In order to carry out the treatment the encrusted metal is allowed 
to soak in solution i, in a covered container of glass, porcelain, or 
stainless steel, either in the cold or, if it is desired to hasten operations, 
in a steam-bath until the incrustation is softened, when most of it can 
be removed by brushing. The colourless Rochelle salt solution be- 
comes deep blue in reacting with the incrustation. 

After the first part of the treatment, a brownish-red residue of 
cuprous oxide (cuprite) will be found firmly adhering to the metal, 
and it does not respond well to brushing. It is usually associated with 
pasty cuprous chloride. Sometimes there is also a layer of metaUic 
copper which has been deposited within the incrustation during the 
process of corrosion. This copper (referred to as redeposited copper) 
tends to seal up the minerals beneath it, and protect them firom attack 
by the reagents. If redeposited copper is observed, this compHcates 
the treatment because it is insoluble in Rochelle salt and in dilute 
sulphuric acid, and it must be removed mechanically. This is not 
always an easy matter and it takes time, but unless all of the copper 
deposit is picked off at this stage, there can be no guarantee that the 
process will give permanent results. 

The object is then placed in solution 2 in a container of glass, 
porcelain, or stainless steel.^ It is set aside in a warm place with occa- 
sional brushing, and examination with a pocket lens, until as much as 
possible of the cuprite has been eliminated together with a sludge 
of copper powder. 

A pecuhar condition that sometimes arises and may be seen under 
the lens is one in which cuprite forms a crystalline mosaic with 
copper crystals on the apparently intact surface of the cleaned metal. 
As this film may have chlorides trapped beneath it, it must be broken 

’ The acid must always be added to the water, and not vice versa; water should never 
be added to strong sulphuric acid as this might cause serious accident due to spurting 
accompanying the sudden evolution of heat. 

^ If a stainless-steel vessel is used with acid, insulate the objea in the vessel by placing 
it upon an old photographic plate or other resistant electrical insulator to avoid deposition 
of the copper on the steel. 

liiiON/i' liusr or 



down in the final operation. It is in helping to break down this film 
that the method of electrolysis is so useful, and the electrolytic tank 
is thus a valuable adjunct to the Rochelle salt method of cleaning. 
Electrolysis is not essential, however, as brushing and pickling in 
dilute sulphuric acid solution can achieve the same result, although 
this usually takes much longer and is more laborious. The object is 
boiled finally in several changes of distilled water, coo ling on each 
occasion in the water to flush the capillaries as described on p. 198, 
The steam-oven method of heating and cooling is more convenient 
when dealing with a batch of objects. Washing is continued until 
chlorides can no longer be detected in the water when the silver 
nitrate test is applied as described on p. 199. 

(ii) Use of alkaline Rochelle salt with hydrogen peroxide. When 
bron2es having fine line ornament are heavily encrusted the sulphuric 
acid, in dissolving the cuprite, tends to leave any fine lines filled with 
copper powder, and in this event the copper is difficult to brush out. 
This comphcation can be avoided by using a single oxidizing bath 
made up by adding 100 ml. of hydrogen peroxide (20 vols.) to each 
htre of the standard Rochelle salt solution i. The object is then 
immersed and brushed from time to time and picked if necessary. 
The action of the hydrogen peroxide may be followed by observing 
the small bubbles that are Uberated at the point of attack, namely, 
around the residues of cuprite; cuprous salts are oxidized and become 
soluble in the Rochelle salt solution. The hydrogen peroxide method 
is slower in action than when sulphuric acid is used, and the metal 
object may be sUghtly attacked in the final stages if the oxidizing 
solution is used for too long. But, under strict supervision, the 
oxidizing process is, on the whole, the more likely to yield a satis- 
factory restoration of fine hne ornament, and it is considered there- 
fore to be the preferable technique where fine decoration exists. 

The beginner tends to seek for a simple course of treatment, and 
may be tempted to ask whether it is not possible to simpHfy opera- 
tions by carrying out the whole of the cleaning with sulphunc acid 
alone. This is unlikely to be successful. Sulphunc acid does not re- 
move redeposited copper, and in fact, owing to the length of time 

B 8157 




required for pickling, using siJphuric acid alone, further copper is 
almost certain to be redeposited in a fine-grained adherent form in 
the neighbourhood of the chloride deposits. This would involve an 
inordinate amount of mechanical cleaning: the redeposited copper 
would have to be removed as otherwise salts might tend to be sealed 
in, with the risk of a further outbreak of corrosion. 

Sulphuric acid is really a clearing reagent for cuprite. It is also useful 
for the pre liminar y treatment of heavy highly-corroded bronzes, 
such as a bunch of spears or tools that have become corroded to- 
gether. In this case it is used at about 20 per cent, strength as a pickle 
to release the corrosion products so that the various objects can be 
separated and then given individual treatment. 

Detailed instructions such as the foregoing tend to suggest that the 
methods are beset with compHcations, but in the appHcation of strip- 
ping processes this is not really the case. Objects do tend to react 
differently under treatment, but this adds interest to the work, and 
at the same time provides a stimulus for ingenuity. 

2. When patina or incrustations should be preserved 

A most difficult problem is presented by the well-patinated bronze 
which is corroding in spots. Here the aim must be to arrest the 
corrosion while retaining the general character of the object. The 
method of treatment is determined in the first instance by the struc- 
ture and appearance of the patina and by the distribution of the spots. 
The following are the three commonest conditions — isolated 
spots in thin soft patina; isolated spots on thick hard patina; and 
spots overlying a porous area. Examples will now be presented to 
illustrate the treatment apphed to typical specimens in each category, 
but methods may have to be varied considerably to suit particular 

An Islamic bowl of sound metal with chased ornament all over it 
had a thin green patina tinged with brown cuprite. The patina was 
spotty (indicating the presence of chlorides) and, moreover, it con- 
cealed important details of the ornament. It was desired to remove 
chlorides, get rid of the spots, and at the same time expose the 



pattem-work where it was concealed, without giving an eSect of 
over-cleaning. The bronze was soaked in changes of sodium sesqui- 
carbonate (5 per cent.) for some weeks, rubbing away the incrusta- 
tion after it had softened in those places where it concealed ornament. 
This could be done with the fingers or with a soft wooden scraper. 
The bowl was washed until chloride was no longer detected in the 
wash-water on testing with silver nitrate, and then boiled in water 
for some hours so as to darken the metal where it had been exposed. 
The bronze was dried and pohshed with a waxed brush and it stiU 
retained the appearance of a patinated bronze of brownish colour 
reheved by green carbonates in the deeper parts of the ornament. 
This type of treatment is often apphed and gives good results with 
thin patinas if they are carefully handled while in the softened con- 
dition but it must be noted that warming the sodium sesquicarbonate 
solution would tend to dissolve the patina. 

A fine Persian bronze vase having a comparatively thick hard 
green patina was similarly infected with a few deep-seated pale green 
spots. These not only detracted ftom the appearance of the vase but 
were observed over some weeks to have increased in size and owing 
to the thickness of the mineralized surface the condition was re- 
garded as serious. In this case the spots were carefully cleaned of all 
powdery matter, using a needle under a binocular magnifier; they 
were then reduced locally, by applying to each individual spot zinc 
powder and drops of 90 per cent, sulphuric acid from a fine glass 
capillary tube. By this means the chlorides responsible for the cor- 
rosion were rendered soluble and could be washed out in changes 
of distilled water. Such treatment, which removes all the powdery 
material and leaves the spots a dull brown colour, is most likely to be 
efiective when, as in the present instance, the general patina is hard 
and shiny, but care must be taken in using acid to prevent it spreading 
as it attacks malachite instantly. After treating a spot and allowing 
effervescence to take place for perhaps half a minute or less, the 
bronze is held in a stream of water, then mopped dry, and the area 
inspected. It may need further excavation and further reduction- 
experience win tell. In restoration work of this kind it is worth while 



malcing experiments on worthless material before attempting to re- 
store an object of value. It should be emphasized that strong sulphuric 
acid is a dangerous chemical (see Appendix II) and is only used in a 
room where a water-supply is available in case of emergency. 

Incidentally, the method when apphed over a larger area provides 
a means for the local treatment of ornament and has been found 
useful in rendering legible incised inscriptions that have become 
clogged with oxide, where it is desirable to avoid wholesale reduc- 
tion. After such treatment, which, of course, removes the green 
incrustation, the metal should be washed locally by a water jet so 
arranged that the washings fall away without running over un- 
treated parts of the object. 

Egyptian bronzes are often heavily infected with salts and for this 
reason seldom possess good patinas. When a good patina does occur, 
therefore, every effort should be made to preserve it, but if the surface 
once becomes spotted, restoration may be a difficult matter. An 
Egyptian bronze cat about lo inches high, having a patina that was 
exceptionally fine, persisted in corroding over a small porous area. 
As there was a natural reluctance to excavate the corroded area, it 
was treated locally with jets of distilled water, arranged to run con- 
tinuously using a water circulator.^ This was continued for many 
weeks in the hope that prolonged washing might be effective. The 
bronze was then dried and the area painted over with lacquer. Since 
this treatment failed to arrest corrosion, the lacquer was removed 
and the washing repeated, but no matter how many times this was 
carried out, nor how many sealing agents were tried, it invariably 
happened that after some months of quiescence the trouble broke out 
anew. It was found necessary in the end to excavate the area where 
the salt was most concentrated and to make good with a httle powder 
colour ground in Ercalene lacquer in order to preserve the appearance 
of the specimen; but such treatment has httle chance of being effec- 
tive unless the object is kept permanently in a dry case, or protected 
by some other means from damp. 

’ Barker, H., and Organ, R. M., ‘A Simple Water Circulator for Museum Use’, 
Studies in Conservation, 1953, i, pp. 84-86. 


The intimate connection between damp and corrosion is iQus- 
trated by the interesting case of a Greek bronze head of a Barbarian 
dating to 400 b.c., which had a very fine patina that had never given 

Fig. 9. Preservation in dry air. The drawer in wooden pedestal is filled with sihca gel so that, 
when closed, the drying agent is in direct contact with the air in the glass case through 

perforations behind the plinth 

any trouble. Suddenly it showed signs of corrosion and the cause was 
traced to damp from a leaky roof that had entered the wall against 
which the bronze had been exhibited. The spots were excavated 
individually with dental tools and, when the bronze had been care- 
fully washed and dried, the cavities were filled with lacquer. No 
further outbreak was detected until the following winter when a 
series of new spots had to be treated, but these were neither so in- 
tense nor so obtrusive as before. There has been no further trouble. 



nor indeed is any expected, as the specimen has now been mounted 
in a case of the island type kept permanently dry by sdica gel (see 
Appendix XI). Such an arrangement is shown diagrammatically in 
Fig. 9. In instances of this kind it is often better to accept the risk of 
further outbreak and be prepared to attend to the specimens again, 
rather than adopt drastic measures that would be certain to change the 
character of the piece. 

It is obvious that the methods already given are capable of being 
varied in many ways to suit the problem in hand. Thus, the water- 
circulator might be used to circulate sodium sesquicarbonate or some 
other solvent, and the local reduction process might be carried out 
using caustic soda in place of sulphuric acid; or aluminium or mag- 
nesium powder might be substituted for zinc. Whatever treat- 
ment may be apphed, the desiccator type of museum case, containing 
sftica gel, is a useful stand-by in helping to arrest any residual activity. 

The foregoing examples illustrate how the choice of method in 
dealing with corroded bronzes is limited by the presence of a good 
patina. It is sometimes limited by the presence of material of a non- 
metaUic nature. Thus, the plaster and gold leaf that covers certain 
types of Egyptian bronze may provide a serious comphcation. Ves- 
tiges of the former ornament may remain, and the bronze cannot, 
therefore, be immersed in any hquid. In one such case, corrosion was 
arrested by daily appHcations of paper pulp (see p. 299) that had been 
soaked in a 5 per cent, solution of sodium sesquicarbonate. Many 
appHcations were made before the spottiness finally disappeared, 
and the metal surface, where exposed, was covered by a stable green 
film of basic copper carbonate. 

Egyptian bronzes sometimes retain a siHceous core as a legacy 
ftom the casting technique, and when such bronzes are diseased it is 
usually a difficult matter to restore their stabihty. As a siHceous core 
is so much more porous than bronze, it contains proportionately more 
salt, and unless the core can be removed, any immersion process is 
Hkely to do more harm than good. Such bronzes can only be given 
local surface treatment. 

The final example is of an Egyptian bronze having a core of a 



different type. This finely chased specimen was a hollow casting of a 
falcon, and the bronze had a deep green patina which, however, 
was pitted and spotty. As various forms of surface treatment had 
been tried without success, it was eventually decided that the salt 
must be inside and that it would be necessary to remove the base- 
plate in order to gain access. When the base-plate was cut out the 
mummified remains of a bird were foimd in the interior, and after 
these had been extracted it was a simple matter to restore the bronze 
by the use of chemical solvents, immersing it first in citric acid and 
finally in sodium sesquicarbonate untd stabihty had been attained. 

3. Removal of calcareous deposits. Use of sodium hexametaphosphate 

Bronzes are sometimes covered with calcareous deposits, and 
while these may be removed with dilute nitric acid this is undesirable 
where there is fine work, or where metal is very thin, as the acid 
treatment is rather drastic. In aU cases it is preferable to remove the 
deposits by soaking the bronze in a 5 per cent, solution of sodium 
hexametaphosphate (Calgon), ^ which in time releases the deposits of 
calcium and magnesium salts by combining with them to form 
soluble salts. Stronger solutions up to 15 per cent, may be used. This 
hastens the reaction, and warming the solution hastens matters still 
further, but not without risk to the patina. On the other hand, where 
the patina is not of importance, Calgon solution can be used to 
remove the green incrustation completely, merely by continuing 
the treatment for long enough. Calgon has the merit that it may 
safely be employed for cleaning the surface of copper alloys that are 
in such an advanced state of decay that no metaUic core remains. For 
this reason it is a useful material to have in the restoration laboratory. 


Chinese bronzes comprise alloys of the gim-metal and bell-metal 
class, containing tin in the proportion of 10-20 per cent. Such alloys 
were used for the massive funerary bronzes. The mirrors are of 
’ Farnsworth, Miss Marie, Technical Studies in the Field of Fine Arts, 1940, 9, p. 21. 



Speculum metal, a white bronze which may contain tin in quantities 
up to about 30 per cent. Lead is commonly present as a minor 
constituent in all Chinese bronzes; its presence seems to have httle 
effect on stabihty, though when present in quantity it tends to lower 
the tone of the patina. 

In aUcahne soils which contain Httle chloride, the corrosion of 
copper and its alloys takes place more slowly and more uniformly 
than when much chloride is present, and the products of the reaction, 
as we have seen, are more Hkely to be stable. They may, in fact, 
constitute a beautiful patina well worth preserving, either for aes- 
thetic reasons or as evidence of antiquity. Such are the patinas found 
on Chinese bronzes. They may be thin and enamel-Hke, moderately 
encrusted, or massive, and they may vary in colour from white or the 
palest turquoise to the positive greens and blues characteristic of 
malachite and azurite; a warmer tone is contributed by the presence 
of cuprite in the lower layers and when exposed at the surface may 
add a colour varying in hue from pale ochre to deep reddish-purple. 
In advanced stages of corrosion the remaining metal may be cracked 
and frail, but the progress of mineralization has not always been so 
destructive, and specimens have survived for 3,000 years that are 
beautifully patinated and can still be described as robust. 

Chinese bronzes seldom require chemical treatment to ensure their 
preservation. They may sometimes be improved in appearance by 
mechanical treatment to remove mud or gross incrustations and 
reveal concealed features of ornament or inscription. Indeed, the 
bronzes that reach Western markets have sometimes a mineral sur- 
face suggesting that a form of mechanical treatment has already been 
appHed to remove incrustations and expose to view the more subtle 
hues of the under layers and, in particular, the pale turquoise layer of 
tin oxide. This oxide, which, in the absence of chloride, is usually 
laid down as a thin coherent film, is normally white, but as it is very 
readily stained with copper compounds, it forms the basis of smooth 
shiny coloured patinas that are stable and much admired and sought 
after by collectors. 

The speculum alloys are characterised by britdeness and this ac- 



counts for some of the features of corrosion pecuhar to the mirrors. 
When the brittle core is unable to adjust itself to the increasing strain 
resulting j&om expansion accompanying the change from the metalhc 
to the mineral condition, cracks begin to make their appearance. 
Sometimes they are superficial and hair-hke (Pi. 32), in which case 
they tend to become centres of corrosion, from which copper com- 
pounds spread out over the adjacent metal without necessarily attack- 
ing it. The rate of decay is slow, however, compared with that found 
in bronzes containing less tin. Sometimes corrosion takes place in iso- 
lated spots, and when this happens in the case of speculum the surface 
of the metal becomes cracked around the spots in a series of small 
concentric rings, so that each spot looks like a tiny oyster-shell, and 
further cracking may even Hft the spots clear of the surface where 
they remain as hard excrescences (Pi. 33). Such pitting is character- 
istic of speculum metal that has at one time been highly pohshed; it 
may be taken as indicating considerable antiquity in the specimen, 
and there is no evidence of any change in the appearance of the spots 
under museum conditions. 


The conservation of copper alloys is often comphcated by the 
presence of decorations in other metals, chief among which are gold, 
silver, and tin, and it may be as important to preserve as much as 
possible of these other metals as to preserve the copper alloy itself. 
Decorations may easily be lost in a heavy incrustation of corroding 
copper and it is very important, therefore, to be on the look-out for 
evidence of ornament, as this might be of great archaeological signi- 
ficance. The discovery of such decoration will obviously have an 
important bearing on the method of treatment that is employed in 

Copper alloys may be gilded with gold leaf or gold amalgam. 
Whichever process has been adopted there is always a problem when 
the metal underlying the gold corrodes, because the gold becomes 
covered with copper compounds so that its presence may not at first 



be recognized. Gold is insoluble in the chemical reagents normally 
employed for cleaning bronzes, but if these reagents should attack 
the corroding bronze lying underneath the gold, there is nothing 
to prevent the gold being lost. If, firom a preliminary examination, 
the presence of gold is suspected, it is preferable to deal with the 
gold first and the bronze, locally if necessary, afterwards. A useful 
reagent for clearing the gold is weak nitric acid (i per cent.), as this 
softens carbonate incrustations and enables the extent of the gilding 
to be assessed. Drops of the acid are apphed with a match-stick to the 
encrusted gold, preferably under a binocular magnifier, and absorbed 
with the dissolved incrustation in filter-paper; to assist the action 
of the acid, any insoluble minerals are cleared away, httle by Httle, 
with the help of a needle. Working in this manner it is possible, in 
time, to expose all of the gilding. As a final operation, the gold may 
be fixed with lacquer after washing and drying. 

Examples are found among Chinese bronzes of the T’ang and 
later periods where the film of gold is coherent and alloyed to soHd 
bronze by the mercury gilding process. In these cases, alkaline 
Rochelle salt may be used as the cleaning-agent. Where there is gold 
embellishment and the underlying base metal has expanded by oxi- 
dation, however, the gold may be porous, and an object in such a 
condition cannot safely be treated with Rochelle salt or, indeed, by 
any other chemical process as any chemical action on the base metal 
would be likely to loosen the gold which would float off. Mechanical 
picking has to be resorted to and is carried out under the microscope, 
particular attention being given to base metal deformations with a 
view to recovering as far as possible the original shape and the 
artistic appearance of the object. 

Silver and tin decorations on bronze survive without corroding, 
but in each case preservation may be diflicult. Reference has aheady 
been made (p. 189) to the cathodic protection afforded to silver by 
contact with copper in presence of an electrolyte, and to the fact that 
a base-silver alloy having a green incrustation may be mistaken for 
bronze and be dealt with as such (p. 220). When white metal is 
observed in treating an object having a green incrustation, the action 



should be arrested by washing the object with water. The partially 
cleaned specimen should then be closely inspected with a magnifying 
glass to determine whether there is any pattern-work before deciding 
what modification of treatment, if any, is necessary in the circum- 
stances. If the white metal proves to be silver (as can be determined 
by its being stained black by a drop of warm ammonium sulphide), ^ 
the probabihty is that warm formic acid, apphed locally to the white 
metal, will be the most suitable treatment, but mechanical methods 
should be exploited as well. When a bronze is embellished with 
silver it is generally desirable to preserve the patina of the bronze and 
in such cases it may be better to switch over to sodium sesquicarbon- 
ate. If the bronze has been reduced, it should be boded in distilled 
water for some hours at the conclusion of the treatment to darken it, 
if necessary, so that the dark metal wdl act as a foil for the silver. 

Should the white metal prove to be tin it is hkely to be in the form 
of a thin coating and in this case reduction methods would be in- 
advisable. Tin is found on bronze cooking-vessels, on ornaments 
such as fibulae and in association with red ‘sealing-wax’ enamels and 
inlays on the escutcheons of Saxon hanging bowls, while, in China, 
it was used for elaborate inlays on Han bronzes — chariot fittings, belt 
fasteners, and sword guards. On all of these objects a heavy incrusta- 
tion of copper minerals is best removed from a film of metaUic tin 
by the application of mechanical methods of treatment — by the use 
of needles working under a binocular microscope. On the other hand 
alkaline Rochelle salt may be made to give satisfactory results but 
there is then no guarantee that all of the tin will be preserved when 
it is attached to metal that has already corroded. 

Extreme caution is necessary when beginning work on freshly 
excavated material. This is well illustrated by some of the objects 
from the ship burial at Sutton Hoo in East Angha, a notable example 
being the large dragon from the shield (Pi. 23) which was 

' A more precise test for silver is to place a drop of 15 per cent, sulphuric acid on the 
metal and a tiny crystal of chromic acid in the drop and wash off after one minute. A red 
stain (silver chromate) left on the metal indicates that silver is present. The stain may be 
easily removed. 



concealed under a green muddy incrustation. The mud was removed 
with the aid of a small stencil brush moistened with a Uttle detergent, 
and the object was found to be gilt-bronze having local areas washed 
with tin, and it was inlaid wdth garnets; in the course of operations it 
was revealed that the tail of the dragon was an ancient restoration 
composed of chalk and glue overlaid with gold leaf Here any treat- 
ment involving immersion in an aqueous solution would have been 
disastrous as the chalk and glue ground would certainly have been 
softened or damaged. 

When dealing with fine metal-work that is decorated with gold, 
silver, or tin, dry or semi-dry methods of cleaning should always be 
appHed in the first instance. There remains the possibility, however, 
that wet methods may ultimately have to be used to ensure perman- 
ent conservation. Also, whatever standard form of treatment is ap- 
phed, it may be necessary to introduce modifications at any stage of 
the work to suit the particular requirements of the specimen. 

The restoration of an Egyptian bronze statuette of Isis and Horus 
(Pi. 34) may be cited as an example of how technique has some- 
times to be modified to meet individual requirements. This bronze 
was heavily encrusted, the white patches showing in the first photo- 
graph being an indication of bronze disease. Prehminary examination 
revealed that the eyes were inlaid with thin sheet gold, but the gold 
was considered to be sufficiently firmly attached to remain unaffected 
by the reduction of the incrustation. 

Treatment was begun by immersing the bronze in a solution of 
alkaline Rochelle salt. The solution was changed frequently with 
intermittent brushing, until the green coating had all disappeared 
and had given place to reddish-brown cuprite. This took three days 
and details of the features were now becoming apparent, as shown 
in the central illustration. Mechanical cleanins; was then carried out 
very thoroughly prior to soaking in sulphuric acid (15 per cent.); the 
acid treatment went smoothly and seemed to remove all the cuprite. 
Thus, a remarkable improvement in appearance was achieved in a 
comparatively short time, but it was clear, having regard to the 
coarse incrustation that had been removed, that the metal must be in 



a highly porous condition and the entire ehmination of salt would 
be difficult. 

The bronze was therefore placed in a nylon bag (as a precaution 
lest the gold should fall from the eyes) and reduced continuously for 
seventy-seven hours in the electrolytic tank using a solution of caustic 
soda as the electrolyte, with periodic brushing. This treatment dis- 
closed the existence of an inscription around the base. The inscription 
was cleaned out as far as possible with a needle, and the bronze was 
then ready for washing. Washing was carried out by the method 
described on p. 197 and had to be continued for six months before 
the final traces of chloride were eliminated. The specimen was dried, 
and the gold on the eyes which had survived perfectly was burnished. 
Although the bronze was now stable and clean, the appearance was 
unsatisfactory, as the metal had a mottled pale straw colour, and this 
made it unattractive as a museum exhibit. This discoloration was 
dealt with by the device of applying warm Plasticine as a coating all 
over the bronze and leaving it there overnight. (Only some brands of 
Plasticine are effective, and these contain traces of sulphur compounds 
which impart just the right degree of tarnish.) The metal had now 
a rich brown-black colour. It was washed with methylated spirit to 
remove grease, and lacquered with Ercalene. Finally a httle chalk was 
rubbed into the inscription so that the characters would be more 
apparent against the darkened surface of the metal. The restoration 
being completed, the object was remounted on its plinth. 

An altogether simpler case was that of the heavily minerahzed 
bronze head shown in Pi. 35. Although swollen and cracked, the 
metal was so substantial that it could be reduced by covering it with 
granulated zinc in an iron basin and boiling for two hours in strong 
caustic soda solution. The bronze was then removed and brushed 
with a steel-wire brush. This revealed ornament that was rendered 
more precise by further reductions in a fresh soda solution. Washing 
proved in this case to be a simple matter, and the specimen was 
eventually dried, and coated with a nitrocellulose lacquer. The 
discovery of cartouches is always interesting, but in this case the 
interest was intensified by the fact that the cartouches revealed by 



the cleaning contained the names of a hitherto unrecorded Ethiopian 
king (Kheperkari Alentamani). The presence of the vulture head- 
dress identifies the head as being that of the queen. 


When thin copper alloys are buried in the ground, they are hable 
to be crushed by the weight of earth that covers them. Before any 
attempt is made to restore the shape, the metal must be softened by 
heating and plunging into cold water. Over-heating will cause ex- 
tensive oxidation of the surface and it is better to soften tittle and 
often during the bending process rather than attempt to achieve too 
much in one single operation. Pis. 36 A and b show how a shield boss 
of interesting shape was restored and mounted. Where there are 
several dents, the order in which they are removed is sometimes of 
importance and the operator works, as a rule, from the main body 
of metal towards the extremities, endeavouring to recover the shape 
with the minimum of movements of the old metal. As for tools — pin 
vices, wooden cramps, clothes-pegs are all useful, but where the 
metal is thin enough, the most sensitive control is that given by the 
hands, the softened metal being pressed against shaped wooden stakes 
held firmly in a vice. Cracks may be reinforced at the back with soft 
solder (p. 286). 

A novel type of problem is presented when only fragments of a 
metal object remain. If, as in the case of the Saxon hanging bowls 
(Pis. 37 and 38), the fragments are large enough to allow the original 
shape to be established with certainty, it may be worth while to 
construct a model from a transparent plastic substance following the 
contours exactly so that the fragments of bronze can be attached to 
it in their appropriate positions. The first requirement is to model 
the bowl in clay in the inverted position and upon this to make a 
strong plaster cast. This cast may then be used as a negative and 
coated internally with a thermoplastic material such as Perspex, a 
technique depending on the fact that sheet Perspex may be softened 
by heat and blown into the required shape by compressed air. Some 



shapes can be made up more satisfactorily in the cold by turning a 
series of Perspex rings on the lathe and building them up from the 
base upwards, the parts being cemented together with Perspex cement 
or with glacial acetic acid. In working Perspex it is difficult to make 
long joints perfectly; when solvents are used, these evaporate and 
there is a tendency for air to be sucked in, thus making the joint 
conspicuous. The most perfect joints are made by firms specializing 
in heat-welding by supersonic diathermy. After the Perspex shape 
has been made, the old metal fragments are easily attached to it 
using either Durofix or chloroform. 

When a large thin bronze vessel is patinated but disfigured by a 
hole, a very effective repair can be carried out by a method devised by 
Murray Pease of the Metropohtan Museum of Art, New York. 
According to this process, a piece of felt is blocked roughly to 
shape, soaked with polyvinyl acetate, and, when partially dry but still 
phable, it is moulded to conform with the required contours. It is 
then fixed inside the vessel and after the adhesive has dried, the out- 
side is made good to match adjacent metal in colour and texture. 




With own With extraneous matter 

degradation (sand, carbon, calcareous 

products material, &c.) 

(minerals) 41, y, 8a 

B. After restoration 


Shape reproduced in Perspex' by che blowing technique 

No treatment. 

!. (a) Warm soapy water; (b) Lissapol froth. 





C ’O 

o > 

y O 

u ^ 

^ 'a, ^ • 

*ts « 53 

. (m u 

:S C .IS - 

o o 










- 2 C B ^ 

O <« o 

3. U ® 

^ *3 '3 

■§ “3 

■' uTS S! u 

- i ” 6 .a 

J3 -C ,y 3 B 


3 3 3 O O 

S OO C/} (J 00 U* 

•c • 




T o 

a. 3 

2 *" o c o 

rS tfl ClO 

60-3 -a 
^7.50 U 

o o 

.. J3 
cx. &, 

O BS 3S 4J 4J 

3m O O b B 
^ j3 j3 3 2 

.**. u (j <1 « 

J J 

” s c £ a 

.o d .3 3 3 
tt ^ j5 ^ ^ 

2 < < w CO 


.y 3 
hi n 

3 u 
.a e 


fl o 
« b 
<j u 



^ Q mS *0 ^ 

w> ^ 

3 V 

u «* 
TJ .2 

« 3 

S « 



C s S 







b 3 
« S' 

> U 

2 s 

s s 
Q to 







ra ^ 
.a ^ 
-3 *2 










O 4 to 






S ; 

9 CL, o ■ 

•5^;§ ■ 

u ^ 53 • 
»- 5 rt 








-3 V ^ 

g 3.S 
60 to 

C^.9 .9 

.£ -c 3 

c G 






« r? T? 
> i-J CU ^ 

b. ^ 

o S 

O M 

3 2 

^ - 3 

m 9 

.iJe S 

^ t" 


Cl* M> 

a* u 

'3 .1^ 
.J3 V 
U J, 


"3 2 


9 c 



N ^ 











G . 

4> ^ 

J « 

S' 3 
1) ^ ' Q 

O rt 60 J 

uJ2 B g 

:a .2 







^ s 









•H <s «n 

B 6167 




Lead and tin are soft white metals, and though very similar in 
appearance, they are easily distinguished by their relative density — 
lead being almost twice as heavy as tin — and by the fact that clean 
lead makes a black mark when rubbed on white paper and tin does 
not. Lead is subject to corrosion in cases where tin is resistant and 
many organic acids which attack lead have no action on tin. Tin 
can be used with perfect safety for coating copper cooking-vessels 
and, in the form of foil, as a hygienic wrapping for foodstuffs. 

Tin becomes oxidized, however, as a result of prolonged burial in 
the ground. In these circumstances the greyish film of stannous oxide 
which initially forms on the metal gradually gives place to white 
stannic oxide. The characteristic corrosion product of lead is the basic 
carbonate which is also white, but the two are readily distinguished 
hy applying a drop of acid which causes the lead incrustation to 
effervesce, whereas the stannic oxide is unaffected. 

Alloys of lead and tin are known as pewter, the hardest variety of 
which contains about 8o per cent, of tin and 20 per cent, of lead. 
Roman pewter contains about 80 per cent, of tin, but the proportions 
of tin and lead in pewter have been found to vary considerably 
throughout the centuries, owing to scrap metal having been so often 
melted down and refashioned. Modem pewters are in a different 
category. Lead is usually absent, its place being taken by antimony 
and possibly some copper, with the result that these alloys are much 
harder and less subject to oxidation than the traditional alloys. 


Museum objects of lead are normally covered with a thin film of 
oxide. When this has grown slowly in pure air, it has a dull grey 



appearance, and acts as a protective patina; when the film of oxide 
has grown in impure air (air that is contaminated with traces of 
organic acids, paint fumes, &c.) it is discontinuous and non-protec- 
tive; a certain milkiness may then make its appearance on the surface, 
and in time active corrosion breaks out with the formation of basic 
lead carbonate. The carbonate is loosely adherent and the corrosion 
is accompanied by a considerable increase in volume. It is for these 
reasons that leaden objects often suffer serious disfigurement unless 
the corrosion is checked in the early stages. 

When leaden objects are excavated firom the ground they are com- 
monly found to have a white incrustation. This is composed of lead 
compounds formed by chemical action between the metal and saline 
matter, and also by the action of oxygen and carbon dioxide dis- 
solved in the ground water. Even when such an incrustation appears 
to be stable, it is often so unsightly as to be unacceptable on specimens 
for study and exhibition. 

The aim of laboratory treatment is therefore either to arrest corro- 
sion, or to improve the appearance of the lead. Carbonates and other 
salts must be removed, and then the cleaned metal is preserved either 
by exposing it in an uncontaminated atmosphere so that it will budd 
up its own protective flhn of oxide (see above), or, more usually, by 
sealing the surface with a film of wax. A serious comphcation arises 
from the fact that the expansion of the surface on corrosion causes 
micro-cracks to form in the underlying metal, and these have to be 
cleared of impurities if the results are to have any permanence. 

Electro-chemical and electrolytic reduction 

While the standard reduction methods apply in the case of lead, 
they are not the safest processes to adopt as lead is sHghtly soluble in 
caustic soda (unless it is cathodically protected) and prolonged immer- 
sion may result in the loss of fine detad. The same type of accident 
might happen during electrolysis if the electric current were switched 
off whde the leaden object was immersed in alkah. Reduction is use- 
ful, nevertheless, for deahng with odd specimens where fine detad is 
absent, and especially with large objects that are heavdy encrusted. 



Special care is required to wash away the last trace of soda from the 
metal, as, if alkah is left in the porous surface or in cracks, the cleaned 
lead will eventually become white. 

Washing lead free from caustic soda. Washing cannot be carried to 
completion by using cold water alone, nor is it possible to be certain 
that the lead is free from alkali unless indicators are used, the most 
useful in this case being thymolphthalein and phenolphthalein. The 
procedure is carried out in two stages as follows: First, place the 
object in changes of very hot tap water^ to remove the bulk of 
the alkah, and test the waters from time to time by adding a few drops 
of thymolphthalein. 2 When the indicator no longer turns the water 
blue this is a sign that the first stage of washing is complete. 
The lead is then passed through a series of baths of hot distilled water 
(that has been freshly boiled to expel dissolved oxygen and carbon 
dioxide) and this is continued until the washings no longer turn pink 
when tested with phenolphthalein. The object is then removed from 
the hot water by tongs (it should not be handled until after waxing), 
mopped dry with a soft clean cloth, and passed through a bath of 
95 per cent, alcohol (industrial methylated spirit). The lead is taken 
from the alcohol bath by tongs, shaken free from Hquid, and placed 
in a bath of molten beeswax, or low-melting (49° C.) paraffin wax 
(which has the merit of giving a less brittle film at room tempera- 
ture), and kept there for a few minutes at a temperature above 
100° C. to remove residual traces of moisture. It is withdrawn from 
the wax, shaken, and placed on blotting-paper to drain. A domestic 
hair-dryer is very useful for keeping wax molten on the object so 
that any unnecessarily thick surface deposits may be wiped off. The 

'■ Waters from peat contain humic acids, &c., and are unsuitable as they tend to dis- 
solve lead. The water from a lead piping system may safely be used for washing. 

* Phenolphthalein is too sensitive for use with tap water, as some drinking-waters are 
faintly alkaline in their action, or develop alkalinity on heating, and in such cases the test 
would not be discriminatory. Thymolphthalein, on the other hand, is not so sensitive to 
faint alkahmty. The indicators remain colourless in acid and neutral solutions, but give 
the following colour reactions in the presence of alkah: 

Thymolphthalein Blue (pH above 9-3 ). 

Phenolphthalein Red (pH above 8-3). 



thin film of wax that remains permeates the porous surface of the 
metal and acts as a protective coating. 

Treatment oj leaden objects with acids 
As the white incrustation that develops when lead corrodes is 
essentially carbonate, it is readily dissolved by acids. 

Dilute nitric acid was formerly recommended for removing in- 
crustations of lead carbonate from antiquities, the cleaning effect 
being immediate. It seemed that excess acid could be neutralized 
with alkali, such as caustic soda, and the soluble residue removed 
with water, leaving a clean metallic surface which, on drying and 
sealing with wax, would remain permanent under museum condi- 
tions. This, however, has not proved to be the case. It appears that, 
after a lapse of several years, lead so treated acquires a irulky appear- 
ance, and the method is no longer considered satisfactory. 

Experiments using acetic acid have given even less permanent 
results judged on a time basis. This is not surprising as lead corrodes 
very readily in presence of acetic acid vapour and carbon dioxide. 
This, indeed, forms the basis of a commercial process for the manu- 
facture of white lead. 

Caley^ has reported favourably on a method of treating lead by 
hydrochloric acid and ammonium acetate, and this is of special inter- 
est as being in the nature of a long-term experiment. Fifiy-six leaden 
objects from the Agora were dealt with in 1937, and seventeen years 
later these were found to be in perfect condition. For this method 
two solutions are required composed as follows: 

Dilute hydrochloric acid: 

100 ml. concentrated hydrochloric acid (Analytical Reagent, Sp. Gr. 
1. 19) in a htre of distilled water. 

Ammonium acetate solution: 

100 gm. ammonium acetate m a htre of distilled water. 

Besides these it is necessary to have a quantity of distilled water that 
has been boiled vigorously a few minutes prior to use (to expel dis- 
solved gases) and then protected from the air. 

' Caley, E. R., Studies in Conservation, 1955, 2, p. 49. 



The procedure is as follows: 

1. Hydrochloric acid bath. Soak object in some fifty times its volume of 

acid till effervescence ceases (time 1—2 hours, or overmght). Drain off 
acid and place object for a few minutes in some hundred times its 
volinne of hot distilled water. Decant and repeat the washing twice. 

2. Ammonium acetate bath. Soak washed object in some twenty-five times 

its volume of warm acetate solution till no more corrosion products 
visible on surface of lead. This should take about i hour and should 
not be prolonged beyond 2 hours at normal room temperatiure. 
Wash as follows: place object in some 100 times its volume of cold 
freshly boiled distilled water. Time ten minutes. Decant and re- 
peat the washing three times. 

3 . Drying. Dry at room temperature without heat, or dry through alcohol 

in the usual way. 

4. Waxing. Immerse the object for a few minutes in molten paraffin 

wax at 100° C. 

The advantage of using ammonium acetate is claimed to be two-fold; 
it dissolves lead dioxide which is insoluble in hydrochloric acid, and 
it acts as a buffer to protect the lead from the action of any trace of 
hydrochloric acid that may remain from the first bath. 

Note: This method is recommended for leaden objects only and 
not for tin, as tin is soluble in hydrochloric acid ! 

Use of ion-exchange resins in the treatment of lead 
A notable contribution to the preservation of leaden objects was 
made in the British Museum Research Laboratory when it was dis- 
covered that lead carbonate and chloride could be removed from 
the metal by the use of ion-exchange resins. ^ 

The ion-exchange principle has long been known as a means of 
softening hard water, i.e. removing dissolved calcium salts which 
form insoluble compounds with soap. It was found that hard water is 
softened by passing it through a column containing certain naturally 
occurring complex sodium sihcates (zeohtes) . More recently special 
synthetic resins were introduced which carried out the exchange 

' Organ, R. M., Museums Journal, 1953, 53, pp. 49-52. 

4 . 



more efficiently. The function of the ion-exchanger is to abstract the 
calcium ions that are in the hard water and replace them with sodium 
ions. When the resin is saturated with calcium ions it can he regener- 
ated by treatment with brine which, in turn, replaces the 
ions with sodium ions so that the resin can be used again. It is claimed 
that the process of regeneration may be repeated indefinitely. 

The above principles may be employed 
in cleaning leaden objects that are en- 
crusted with carbonate. The process is 
very simple. All that is necessary is to 
place the corroded leaden objects in 
contact with granules of an appropriate 
ion-exchange resin, ^ cover with distilled 
water and keep hot, changing the resin if 
necessary until the white incrustation 
disappears. MetaUic lead is unaffected, 
but the incrustation of basic lead carbon- 
ate which gradually dissolves is removed 
from the system; the lead ions are taken 
up by the resin in exchange for hydro- 
gen, and carbon dioxide is evolved from 
the hot hquid. The ion-exchange method 
is a great advance on previous methods 
that have been used for the treatment of 
lead, as no chemicals have been introduced 
into the system, and therefore subsequent 
washing of the lead is unnecessary. Although the initial outlay m 
resin is rather expensive, the method is economical because the resin 
may be regenerated after use by treating it with nitric acid to dissolve 
out the absorbed lead, and washing it with distilled water till the 
washings are neutral to htmus paper: the resin is then in a condition 
to be used again. The apparatus is shown diagrammatically in Fig. 10 
and the column used in regenerating the resin may be seen (middle 
left) in the general view of the Laboratory, Pi. 39. The method has 
‘ ‘Amberlite IR 120’, British Drug Houses Ltd., Poole, Dorset. 

RUHOt aunc 


Fig. 10. 

Treatment of a corroded 
object by the ion-ex- 
change method 



the further advantage that, when specimens are badly swollen and the 
cracks filled with a resistant white deposit of basic carbonate, it is not 
essential to remove this, as in using ion-exchange resins the carbonate 
has not been contaminated with chemicals. Nevertheless it will 
usually be desirable, for the sake of appearance, to remove all of the 
carbonate if possible. 

In applying the ion-exchange method the following points should 
be noted. Ion-exchange can only be conducted in the presence of 
water, and as distilled water has a certain solvent action on lead, the 
rate of cleaning should be speeded up whenever possible. To facih- 
tate this, any hard incrustations of carbonate occurring in cracks 
should be opened up by picking with a needle. There is httle likeU- 
hood of serious damage by water, however; at the worst there may 
be a tendency for ornament in rehef to be rounded, but the effect is 
very sHght and could only be of significance in cleaning leaden coins 
where the inscription is barely legible. 

The use of ion-exchange resins is advocated for the treatment of 
small objects — coins, medals, weights, badges, sling stones, &c., but 
when the objects are heavily corroded, or when they have been 
treated with shellac or other varnish, it has been found advantageous 
to give them a preliminary treatment for a short time in the electro- 
lytic tank. 

Storage of leaden objects 

The susceptibihty of lead to attack by organic acids is one of its 
principal characteristics. The effect of carbonic acid in presence of 
acetic acid vapours and of humic acids has already been noted. Tannic 
acid is equally corrosive, and an exudation of tannic acid from oak 
cupboards and drawers is credited with causing the intensive corro- 
sion to which leaden objects are subject when stored therein. Coins 
and medals, beggars’ badges, leaden crosses, inscribed leaden scrolls, 
&c., have been found converted completely to amorphous masses of 
white powder as a result of their being kept during past generations 
in oak cupboards. Nor are other woods necessarily innocuous. To be 
harmless, the timber would require to be well seasoned, and if any 


doubt exists, it should be sealed with a resistant lacquer. This apphes 
particularly to nests of shallow drawers where there is a large surface 
of wood and limited ventilation (Pi. 40). WeU-seasoned mahogany 
is probably the safest wood to use for cupboards and cases designed 
for the storage of leaden objects. 


Tin has been described as being very stable under normal atmo- 
spheric conditions, but when buned in the ground and exposed to 
moisture and oxygen over a long period it loses its lustre, becoming 
granular and grey in appearance. This change is due, in the main, to 
oxidation. The oxide layer may have a greenish hue when traces of 
copper are present. A more advanced stage in the corrosion is the 
conversion of the grey stannous oxide to a higher form, stannic oxide, 
which is nearly white in colour, hut it is probable that this only occurs 
as a secondary reaction in the presence of soluble salts. 

Tin is often found as a thin silvery layer on bronze, especially in 
jewellery and suchlike. This has been apphed to the bronze in much 
the same way as a plumber apphes soft solder. The bronze has been 
cleaned and fluxed to prevent it oxidizing, heated to a temperamre 
shghtly above the melting-point of tin (232° C.), and the molten 
white metal then wiped over it with a rag charged with tallow. Thin 
washes of tin are easily destroyed both by acids and alkahs, and when 
found on ancient metal objects these should be cleaned by mechanical 
methods if possible. 

Sohd objects of tin, e.g. plates, jugs, coins, and medals, are cleaned 
either electro-chemicaUy or by electrolytic reduction in the usual 
caustic soda electrolyte. Corroded tin coins have been fovuid to 
respond weU to electro-chemical reduction in the cold, using either 
zinc, aluminium, or magnesium powder and caustic soda. 

The problem of corroded tin is often confused with an allotropic 
change that takes place in the metal at low temperatures, known as 
tin pest’. This is, however, a different phenomenon to that of metalhc 
corrosion. It is speedier in action, and more catastrophic in its results. 



for the metal changes its crystalline state and falls into a coarse grey 
powder. The destruction of everything of archaeological interest is 
thus complete, and restoration is impossible. It is fortunate that true 
‘tin pest’ occurs very rarely, and only in exceptional circumstances 
has it been identified with any certainty as occurring in the case of 
museum objects. When an object of tin shows signs of decay these 
are almost invariably the result of metaUic corrosion, and are there- 
fore amenable to treatment. 

Special care is required in dealing with heavily oxidized tin that 
bears incised ornament because in such cases the detail may be en- 
tirely in the oxide layer, which must therefore be preserved, and, of 
course, reduction would destroy this. A case in point is the Hartogs 
plate in the Rijksmuseum, Amsterdam. This famous plate was re- 
covered after more than eighty years of exposure on the coast of 
Austraha, and it had been supposed to be a victim of ‘tin pest’ be- 
cause of its fiail condition. It was shown, however, by micrographic 
examination that the weakness had resulted simply from prolonged 
metallic corrosion. Oxidation of the surface had been accompanied 
by a considerable increase in volume, and this had resulted in strains 
that tended to make the oxide layers split away from the core of 
metal, threatening the loss of the inscription. In such cases, reduction 
is out of the question, and the only possible treatment is to consoHdate 
with a suitable adhesive such as Durofix. Fortunately, in the case of 
the Hartogs plate, there was no indication that corrosion was still 
taking place, and chemical treatment was therefore not required. 

In the case of tin coins that have been struck and bear a raised 
inscription, however, conditions are different. During corrosion the 
inscription is preserved at the expense of the worked metal consti- 
tuting the flat part of the coin. This is an interesting case of the 
debasement of a metal in those parts that have been subjected to the 
greatest mechanical strain. Thus in struck coins the inscription may 
be clarified by careful reduction as was done in the case of corroded 
tin coins from Malaya^ where good results were obtained by using 
magnesium powder and caustic soda. 

' Plenderleith, H. J., and Organ, R. M., Studies in Conservation, 1953, I, pp. 63-72. 


Efiibedding lead and tin in plastic materials 
When a lead or tin coin is of unique interest but is so weak or 
fragmentary that the remaining pattern is likely to succumb to any 
process of cleaning, the only way of saving it is to embed it in 
a transparent plastic. This can be done with Marco Resini S.B. 26 C. 
which sets at room temperature without the use of pressure. Purves 
and Martin^ have described the method of embedding biological 
material in this resin, and Miss Plesters^ has given full details of its use 
for embedding paint fragments that are to be prepared as cross- 
sections. When it is desired to preserve coin fragments, the most con- 
venient shape of the polymerized resin block is one that may be kept 
in a coin cabinet, i.e. a disk, which should be highly poHshed to offer 
minimum interference with vision. As there is a certain shrinkage 
when the resin sets, it is very difficult to prepare such a poHshed disk 
by simply allowing the resin to set in a mould with pohshed plain 
faces. Hence it is usually necessary to work and polish at least one of 
the flat faces. This may be a lengthy operation. The Marco resin pro- 
cess is unsuitable where copper is present, as copper produces a green 
stain in contact with the resin. 

If an electrically heated press is available, a thermoplastic resin 
may be used, and the embedding operation is greatly facflitated, 
since blocks may be prepared of uniform size that only require 
final poHshing. A methacrylate resin called Transoptic'* has been found 
to give reproducible results under these conditions. The powdered 
plastic should first be spread on white paper and examined carefully, 
any dust or foreign matter being removed. The object is then washed 
in chloroform, air dried, and thoroughly dried over silica gel. (Avoid 
touching with the fingers after washing with chloroform.) A circular 
mould, I J inches in diameter, is then partly filled with the powder 
and the object placed gently on top. An equal amount of powder is 
added and the plunger of the mould inserted. The mould is then 

' Scott, Bader & Co. Ltd., 109 Kingsway, London, W.C. 2. 

^ Purves, P. E., and Martin, R. S. J., Museums Journal, 1949, 49, pp. 293-6. 

^ Plesters, Miss Joyce, Museums Journal, 1954, 54 > pp- 97-101. 

* Messrs. A. GtJlenkamp & Co. Ltd., 19 Sun Street, London, E.C. 2. 



placed in the press and heated gradually to 150“ C. without appUed 
pressure. Heating is discontinued, the pressure raised to 2,000 lb. per 
sq. inch, and the mould cooled to 40° C. while maintaining the 
pressure. Cooling may be hastened by a water-cooled jacket. The 
clear resin block containing the coin is then removed and may be 
finished off by polishing with I.C.I. Perspex Polish, grades i, 2, and 3 
being used successively on Selvyt cloths. 

If at a later date it is desired to remove the embedded object, the 
block is soaked in chloroform. Transoptic swells and dissolves; Marco 
resin is crazed, and in each case the object may easily be removed 
without any risks of damage. The thinnest lead foil has been mounted 
in Transoptic and demounted without distortion. 


Pewter vessels depend to a large extent upon their patina to give 
them an appearance of dignity and maturity; chemically cleaned 
pewter has the appearance of lead and so in the conservation of 
pewter objects it is essential to avoid overcleaning. Where the metal 
is merely stained or covered with a film of oxide and it is a question 
of improving the appearance, a rub with a rmld abrasive such as 
rottenstone applied with an ody rag will be all that is required. 

When pewter carries a fine design hidden by a thin deposit of lead 
carbonate, as in the case of the BastiUe medal (Pis. 41 A and b), the best 
treatment to adopt is, without doubt, the ion-exchange process 
described (p. 263). This removes the corrosion product while pre- 
serving the finest details of the design. 

Old pewter is occasionally found to have wart-hke growths of 
incrustation on the surface, possibly arising as a result of localized 
contamination with salts. If these spots have a hard skin and are not 
showing signs of active corrosion, it may be safer to leave them alone 
than to attempt to get rid of them (e.g. by grinding) because the 
material underneath may be mainly tin oxide which is often very 
crumbly. If exposed to the air, moreover, the under-layers of such 
spots may begin to corrode actively. If corrosion is already active and 



treatment essential, electrolytic reduction may be carried out. This 
will remove all corrosion products but the appearance of the object 
will suffer because the surface wiU be left full of small depressions or 
pock-marks. In such a case the only course that can be recommended 
is to fill the depressions and make good the surface with wax. 

Reduction is the safest procedure, however, when searching for 
the original pewter marks imder a layer of heavy incrustation. Such 
marks are easily damaged by scraping as the underlying metal may 
be softer than the incrustation. Tankards and flagons are often marked 
beneath the hp, and care should he taken with capacity marks on the 
side of vessels, and makers’ and owners’ marks under the base. When 
the site of the mark is known or suspected, local reduction with zinc 
and caustic soda may be all that is required (cf. p. 193). 

An example of the unexpected discovery of a pewter mark oc- 
curred in the case of a plate brought up from the sunken galleon in 
Tobermory Bay. This had a thin greyish-white incrustation and was 
apparently devoid of documentary interest. A short reduction in the 
electrolytic tank, however, revealed a stamp on the under surface of 
the plate, and it was sufficiently legible to be recognized by the Duke 
of Argyll as indicating that the plate was of Portuguese origin. 























•3 “ 

s o 

m 'O 





3 r2 

tC fsT 





^ o 






Ih •<> 

— O -a « 



TS v% 






"S o 
2 ^ 







~ ffl ” 

« o 



< 2 







J3 -d 

S ^ 5P 

s -il 
2 I 

** ”S V 
Z ^ £, 

fS v-» 

'o r^ 00 

Solvents: Hydrochloric acid treatment followed by ammoniunt acetate. 

Reduction; (a) (*■) Electrolysis in caustic soda. (Lead being soluble in caustic soda, these processes are applied for short 

Protective finishes: (a) Paraffin wax dip; (b) Impregnation with a plastic substance or embedding. 

Mechanical: (a) picking, (b) chipping, (c) grinding, (<0 scratch-brushing, (e) grit-spraying, (/) and burnishing. 

Storage and exhibition. Avoid the use of oak furniture or unseasoned wood. 



Iron occurs in the metallic condition in meteorites where it is asso- 
ciated with small quantities of nickel, cobalt, copper, &c., and the 
oldest iron objects that have been subjected to chemical analysis have 
been shown by the presence of these trace elements to be of meteoric 
origin. Terrestrial iron is of rare occurrence as the metal is so readily 
oxidized and converted to minerals, and these are abundantly distri- 
buted throughout the earth’s crust. 

Objects of iron and steel provide some of the most intractable 
problems for the conservator, because of the variety and complexity 
of their corrosion products. Iron corrodes easily, the corrosion pro- 
ducts are unsighdy, and the sweUing and deformation of the objects 
may be severe. 

Rusting. Iron is readily attacked by oxygen in presence of moisture 
to form rust — a name derived from the characteristic orange and 
red compoimds that appear as the first products of corrosion. These 
consist at first of a mixture of ferrous and ferric hydroxides, but on 
further oxidation the rust becomes substantially a hydrated ferric 
oxide, in which some carbonate is usually present as well. 

This might seem to be fairly straightforward, but when salts are 
present that can act as electrolytes, the chemical reactions that cause 
corrosion are reinforced by electro-chemical reactions, and minerah- 
zation is greatly accelerated. It was noted in introducing the subject 
of corrosion (p. 190) that when a metal was partially protected from 
aeration by patches of oxide, &c., certain areas became anodic and 
others cathodic so that, in fact, the corroding metal behaved in the 
presence of an electrolyte as if it were a number of tiny galvamc 
cells. This is a feature of iron corroding in the presence of sodium 
chloride, the anodic areas dissolving to form ferrous chloride whilst 



the cathodic areas become alkaline due to the formation of sodium 
hydroxide. A film of hydrogen gas begins to accumulate on the 
cathodic areas, and as this has a comparatively high resistance to the 
passage of electric currents, its presence tends to slow down the re- 
action. In presence of oxygen, however, the hydrogen is continuously 
removed, the two gases reacting to form water or hydrogen peroxide. 
Thus the electrolytic action is enabled to proceed and it continues 
until the area is covered with a deposit of rust formed by the inter- 
action of the ferrous chloride and sodium hydroxide. New areas then 
take on the roles of anode and cathode and since sodium chloride is 
regenerated at the same time as the rust is being deposited the cycle 
of reactions can continue. This is an important factor in the conser- 
vation of iron as it means that we are faced with the same conditions 
that confront us in the treatment of ‘bronze disease’, namely, that 
stabihty cannot be assured until all the chloride has been removed 
from the corroding object. 

It has been mentioned that the accumulation of a film of cathodic 
hydrogen slows down the electro-chemical action, but that the action 
of oxygen in breaking down this film makes it possible for the cor- 
rosion to proceed. It does not follow, however, that, in the absence 
of oxygen, iron will necessarily remain uncorroded. The film may 
be broken down in another way, namely by bacterial action, and 
this introduces a novel factor that is important in studying the cor- 
rosion of ferrous metals. Iron and steel are often found to be very 
badly corroded when buried, under anaerobic conditions, in heavy 
clay containing sulphates. In this case corrosion has been traced to the 
presence of sulphate-reducing bacteria which act in a two- fold 
manner: firstly, converting the sulphates to sulphides which attack 
the iron, and secondly, breaking down the hydrogen film and thus 
making it possible for corrosion to continue. When corrosion takes 
place under such circumstances the surface of the iron is found to be 
covered with a black crust of iron sulphide, and the clay surrounding 
the object is also stained black. Should there be any doubt about the 
nature of the deposit, iron sulphide is easily detected by its reaction 
with acids to hberate sulphuretted hydrogen recognizable by its 

40 . pilgrims’ badgls, blllae, cloth marks, etc., of lead 

These show various stages of disintegration cliaractcristic of the change from metallic lead 

to a basic lead carbonate 



characteristic odour of rotten eggs. The first of the sulphate-reducing 
bacteria to he made the subject of laboratory experiment was Vibrio 
desulphuricans, but many new varieties have since been isolated and 
it is probable that there are several species. ^ This type of corrosion is 
very prevalent and is a common cause of the destruction of iron 
pipes buried in clay. 

Reference should also be made to the action of sheath-forming 
bacteria (e.g. Gallionella ferriginea) in promoting corrosion by differ- 
ential aeration.^ These bacteria form a blistered structure of tubercles 
that is a common feature of heavily corroded iron, and sulphate- 
reducing bacteria may also be active in the anaerobic areas beneath 
this cellular structure. 


When an iron object comes to the laboratory for treatment, a 
careful examination is necessary to determine the condition of the 
specimen before a course of action can be decided upon. If a massive 
incrustation of dry rust is free from chloride, it may be stable under 
museum conditions, or if rusting has proceeded to the limit and no 
metal core remains, even if the oxide is swollen and fissured, the 
specimen will also have reached a stage of stability and no treatment 
will be essential. On the other hand, if corrosion is stiU taking place, 
some process for the removal of chlorides will have to be employed 
in order to preserve the object. Therefore the first question to decide 
is whether there are any signs of chemical activity. 

An examination is first made of the surface. Corroding iron does 
not display the range of colour that is apparent in bronzes that are 
‘diseased’, but even so areas of active corrosion may often be detected 
by the fact that they have a sHghtly different colour and texture to 
those of the surrounding metal or incrustation. But changes of colour 

• For an interesting resume of this work see Butlin, K. R., ‘Bacteria that Destroy 
Concrete and Steel’, Discovery, 1948, pp. 151-5. The general principles of metaUic cor- 
rosion arc broadly surveyed by Vernon, W. H. J., in Research, 1932, 5, pp. 54-61. 

* Burgess, S. G., ‘Chemistry of Water Conditioning’, School Science Review, 1952, 34, 
P- 175- 

B 6157 




and texture do not necessarily indicate in themselves that corrosion 
is active. Dampness on the surface of the object is, however, a sure 
sign, because when chloride is present in contact with iron it gives 
rise to corrosion products that are hygroscopic and beads of brown 
hquid may he found on the surface of otherwise dry rust; even a 
comparatively thin film of rust wiU show signs of dampness due to 
chemical activity if chloride is present. 

For assessing the internal condition of an iron or steel object, the 
method of X-radiography is particularly useful. Iron oxides are rela- 
tively much more transparent to X-rays than the soHd metal, and 
X-radiographs reveal the distribution and extent of oxidation more 
certainly and directly than any other method; they enable one to 
estimate the depth of pitting, and hence to decide on the best method 
of treatment. 

But only in rare cases can radiographic methods of examination 
be considered an essential and their apphcation in any case is limited 
by the thickness of the metal. Whether facilities are available for 
X-radiography or not, a careful examination is always carried out, 
first with a magnet in order to determine the extent of the metal core, 
and then by a needle or metal probe, used with the aid of a lens, in 
order to reveal the extent and condition of the rust layers and dis- 
cover, if possible, the existence of any decoration hidden beneath the 
incrustation. Other tests are made instinctively, e.g. the balance of a 
rusty blade as an indication of the regularity or otherwise of the 
corrosion, and an estimate of weight for bulk (relative density) 
which, with the magnet test, should help in determining the extent 
of oxidation. 


From the preliminary examination it will be decided in which 
category to place the specimens, and guidance as to subsequent pro- 
cedure may then be obtained by consulting Tables VI and VII 
below (pp. 282 and 290). 

Where a substantial core of metal remains in the heart of the rust, 
reduction methods can be employed without the cleaned object being 



mechanically weakened, but if, as in the case of a thin and heavily 
rusted blade, the metalHc core is discontinuous, reduction methods 
are best avoided. 

When a small object no longer has a metallic core it is very brittle, 
and as any treatment would be hkely to weaken the specimen stiU 
further, objects in this category are best left alone. Sometimes inter- 
esting ornament may be hidden in a mass of rust, and can never be 
revealed by cleaning because of the frailty of the specimen. In such 
cases photographic evidence may be made available by radiographic 
methods, using X-rays or radio-isotopes. * At other times a httle 
careful work with a needle under a binocular microscope may suc- 
ceed in reveahng something of special interest. When this can be done 
without imdue risk it is worth while in specimens that are designed 
for exhibition in pubhc galleries. 

Laboratory attention may not be required in order to ensure the 
stability of a specimen, yet it may be desirable in order to improve 
the general appearance. Apart from the matter of ordinary repairs, 
treatment may be required for any of the following reasons — to 
reveal the original contours or shape; to expose ornament existing 
in the rust, or decorative inlays that are hidden by the incrustation; 
to study the method of construction, or, in general, to reveal evidence 
which, in the absence of such treatment, would be hkely to remain 
concealed. Such evidence may sometimes be obtained by radiography, 
but more generally by the apphcation of mechanical and chemical 
methods used either alone or together. Mechanical methods are of 
special significance in dealing with corroded iron and steel. 

Reduction. The standard practice Avith iron objects is to apply re- 
duction methods whenever possible, i.e. when a good continuous 
core of metal remains, and there are no comphcations in the way of 
inlays or non-metaUic attachments. In cases where the surface is not 
much pitted, electrolytic reduction may be apphed forthwith, but 
when pitting is at all extensive it is advantageous to start with electro- 

• Moss, A. A., ‘The Application of X-rays, Gamma Rays, Ultra-Violet and Infra-red 
rays to the Study of Antiquities’, Handbook for Museum Curators, Part B, Section 4, 
Museums Association, London, 1954. 



chemical reduction, using zinc and caustic soda, as this treatment is 
more likely to get at the deep-seated corrosion in the pits and cracks; 
but a final electrolytic reduction is always worth while, as it has been 
found to remove most of the residual chloride, and thus cut short the 
time necessary for washing. This is clearly an advantage in dealing 
with metals that rust. 

Difficulty is sometimes found in making a good electrical contact 
with corroded iron objects, and it may be necessary, by scraping 
away the rust, to expose some uncorroded metal for this purpose. 
Electrical contacts may have to be made in several places in preparing 
large objects for electrolysis. 

As a matter of interest, electrically reduced iron often steams when 
first rinsed after removal firom the tank. This is due to the rapid oxida- 
tion of pyrophoric iron, and a case is on record where a large piece 
of black reduced rust, when picked off, actually glowed and burned 
through the French polish on the bench! This, however, is very 

Caustic soda treatment. Where chemical corrosion is observed to be 
taking place on an object that is too frail for electrolytic reduction, it 
may be treated by boiling the object in several lots of dilute caustic 
soda solution followed by boihng in changes of distilled water. The 
results may be satisfactory when the treated object is kept in a dry 
place, but with this method there is a chance that chlorides may 
remain in the under layers, and the result of caustic soda treatment 
is therefore not necessarily permanent. 

Heat treatment. The rough and ready method of separating a mass 
of corroded iron objects by the use of a blow-lamp is obviously 
attended by the danger that specimens may fly to pieces. This is par- 
ticularly hable to happen in the case of an object having a bubbly 
surface of ferric oxide in which air and salts may be trapped. When 
iron is damp or badly rusted, heating is hable to cause a good deal of 
flaking from the surface, and although it loosens deposits of car- 
bonate, it hardens any clay deposits and makes the rust impervious 
to the action of solvents. The method, therefore, cannot be recom- 
mended without serious reservations. 



Use of rust softeners, solvents, and inhibitors. Thin rust spots on bur- 
nished steel may often be softened by keeping the steel under paraffin 
oil for a few hours, and then rubbing locally with worn emery paper 
of fine grade or a match-stick, but care must be taken to remove all 
the paraffin oil or fresh rust is likely to be formed. The paraffin oil 
is removed with a dry cloth, and replaced with a lubricating oil, and 
this is eventually removed by dry rubbing, any thin residual film of 
lubricating oil being harmless. The substance known as ‘Plus-Gas 
Fluid A’l is even more effective in softening rust than oil. Unless rust 
spots are removed, they tend in time to bite deeply into the metal. 
At these points anodic attack is taking place, and this localized 
chemical action is a stimulus to greater activity. It is important, there- 
fore, in maintaining collections of armour and the like, to be aware 
of the danger of rust-pitting, and to take action in good time to 
remove the rust. Where salts are absent, and atmospheric relative 
humidities are at 50 per cent, or under, rusting is not a serious 
problem, but moisture readily condenses on cold metal surfaces, and 
condensation is always a menace if there should be a sudden drop in 
temperature. Also in a city atmosphere, condensed moisture might 
well contain sulphur dioxide which would promote rusting. A smear 
of lanolin (wool fat)^ has proved to be a good protective agent for 
burnished steel, but as it leaves the surface of the metal sticky, it is 
only of practical value when the steel is protected by being enclosed 
in an exhibition case. A thin layer of non-sticky wax is to be pre- 
ferred for museum objects (see Appendix XII). 

Certain commercial rust removers come within the category of 
rust solvents,^ and these are of special value for derusting objects that 
are either too heavy or too large and unwieldy for treatment by 
electrolytic methods. Some have the added advantage of being rust 
inhibitors, i.e. they leave a protective film on the surface of the metal 
after treatment. Rust solvents may be apphed by brushing, and when 

' The Plus-Gas Co. Ltd., 89 Cromwell Rd., London, S.W. 7. 

^ Lanolin Rust Preventers, D.S.I.R., England, Special Report No. 12. 

^ e.g. ‘ACP Deoxidine No, 125’, Imperial Chemical Industries Ltd., Paint Division, 
Slough, Bucks., and jenohte’, Jcnohte Ltd., 43 Piazza Chambers, Covent Garden, 
London, W.C. 2. 



the solvent action has gone far enough, either the surface is wiped 
clean without washing, or in other cases any remaining chemicals are 
washed away. In using these commercial preparations the makers’ 
instructions must be followed impHcitly. 

In these rust inhibitors the active constituent is often a derivative 
of phosphoric acid which forms an inert film of ferric phosphate on 
the surface of the metal. In this connexion it is interesting to observe 
that objects of iron or steel are occasionally found in the ground in a 
substantially uncorroded state after years of burial. This may some- 
times be explained when they are from a burial site, and have been 
exposed to the action of phosphates from adjacent bones. The pro- 
tection given by tannates, which may have their origin in decom- 
posing leather or oak bark, is even more marked. 

It happens that caustic soda is also a rust inhibitor (p. 276) , and this 
property gives it an added advantage over acid electrolytes in dealing 
with ferrous metals. The metal objects must still be washed after re- 
duction to remove dirt and the last traces of chloride, but if the 
final wash water is rendered faintly alkaline with caustic soda (and 
the object dried thereafter), rusting will be inhibited. 

Rust may be removed by the use of a solvent such as a 9 per cent, 
solution of oxahc acid, but it is often more convenient to use a com- 
plexing reagent, which has the property of combining with the iron 
dissolved from the rust; the iron is then said to be sequestrated. Some 
of the most effective complexing reagents are derivatives of ethy- 
lenediamine tetra-acetic acid (EDTA products), and examples are 
the Versenes,^ a range of compoimds that will complex iron under 
acid or under alkaline conditions. ‘Versene Fe-3 Specific’ is claimed 
to be the most powerful complexing agent for dealing with iron salts, 
but Versene T is of greatest interest as it can be used in presence of 
caustic soda which slightly dissolves rust to form ferrites. This is 
extremely effective on freshly deposited rust, but less so on aged and 
especially on heated iron oxide. In these difficult cases it is a help to 

* Berswoith Chem. Co., Framingham, Mass., U.S.A. Manufactured by Rexolin- 
fabriken, A. B., Halsingborg, Sweden, Distribution in England by F. W. Berk & Co. 
Ltd., Commonwealth House, 1-19 New Oxford St., London, W.C. i. 



add sodium hyposulphite.* A solution which has been used with 
effect contained 30 ml. ‘Versene T’, 30 gm. caustic soda, and 6 gm. 
sodium hyposulphite per htre. The method of appHcation is to keep 
the rusty object immersed in a hot solution until such time as the 
rust is dissolved to the required extent, the object being thoroughly 
washed thereafter and dried. 

Protective finishes. Whatever chemical method is adopted in clean- 
ing iron and steel, the subsequent washing is likely to be accom- 
panied by a certain amoimt of rust discoloration, and when the metal 
is dry this is removed by brushing, but if washing has not been 
prolonged, the rust discoloration can be avoided by a final bath in 
acetone which removes the surface film of water. There then arises 
the question of what type of protective finish is likely to be most 
suitable. Choice may be made firom a wide range of materials: there 
are sealing agents, such as oils and fats, waxes or lacquers: there are 
also chemical inhibitors, one of the most effective of which is sodium 
benzoate. This may be appUed in either a water or glycerine solution 
of 1*5 per cent, strength for keeping steel bright in moist sur- 
roundings. See also Appendix XII. 

Certain chemical substances have the property of preventing the 
corrosion of iron and steel in enclosed spaces without being in con- 
tact with the metal surface at all. These are known as Vapour Phase 
Inhibitors (V.P.I.),^ but it is not safe to use them indiscriminately, as, 
however effective they may be with ferrous metal, some have been 
found to attack copper alloys and others soft solder. If such defi- 
ciencies can be overcome, as seems not impossible, vapour phase 
inhibitors should have wide apphcation in the museum world in 
helping to preserve objects of iron and steel, but meantime it is 
likely that the use of wax or lacquer will continue to be the most 
popular type of finish for archaeological specimens. In some com- 
parative tests conducted at the Ancient Monuments Laboratory,^ 

' Formerly known as hydrosulphite. This is not to be confused with photographic 
‘hypo’, which is sodium thiosulphate. 

^ Biek, L., Museums Journal, 1953, 53, pp. iio-ii. For summary of tests see Bennister, 
H. L., Research, 1952, 9, p. 424. 

* Biek, L., Cripps, E. S., and Thacker, D. M. D., Museums Journal, 1954, 54, pp. 32-36. 

28 o 


dipping in molten bleached beeswax was shown to be the most 
effective method of preserving treated objects. 


A feature of the Japanese sword-blade is the pattern latent in the 
steel, arising from the method of forging by folding and twisting 
the red-hot billet of metal according to a pre-arranged formula. A 
series of intricate cloud, wave, and other patterns may thus he re- 
peated along the length of the blade, and good examples are greatly 
esteemed by the connoisseur. Should the metal he allowed to become 
dull the pattern may become indistinct and finally disappear. Refer- 
ence has heen made (p. 206) to the conservation of such weapons. 
The following note refers to a method of restoring a lost pattern, 
using the reagent called ‘nital’ prepared by dissolving 1-5 ml. of con- 
centrated nitric acid (Analytical Reagent, Sp. Gr. i. 42) in 100 ml. 
of alcohol (industrial methylated spirit). First remove any grease with 
a soft rag moistened with benzene. Moisten a swab of surgical cotton- 
wool with freshly prepared nital solution and apply it uniformly to 
the blade, rubbing backwards and forwards from hilt to point. This is 
done beside a running tap under which the blade is plunged to arrest 
the etching of the acid. Sometimes this solution develops the pattern 
within a few seconds; at other times several short apphcations may 
be necessary before the desired result is obtained. The blade should 
be closely inspected after each washing; as an excessive use of the 
reagent might blurr the pattern, it is very important to avoid carrying 
the action too far. The blade must finally be washed imder the mn - 
ning tap for about half a minute and carefully dried with a soft cloth. 

Japanese swords should be kept slightly waxed. If the waxing is 
overdone, there is a possibdity that dust will accumulate, and as great 
care must be taken to ensure that the scabbards are kept free from 
any dust that might cause scratching, it is better not to wax at all 
than to apply wax too Hberally. The formula using microcrystalline 
wax given in Appendix XII is suitable. The smallest quantity is 
brushed uniformly over the burnished metal with a moderately 



Stiff brush such as is used for cleaning silver. Then as much as 
possible of this is removed by brushing with a softer brush having 
longer bristles — the kind of brush that is used for silk hats. 

The textured patterns in damascened blades often have the appear- 
ance of ‘watered sdk’, an effect obtamed by long heating and hght 
forging, producing what is known as spheroidization of iron carbide 
in a pearhte structure. Damascened blades should be treated with the 
same care as is afforded to Japanese swords. 

The coarser pattern effects sometimes found in swords of the 
eighteenth century, and typically in the Malayan kris, are obtained 
by variations in rehef of the different elements of the pattern. This is 
attained by an etching process and in general such blades provide no 
compHcation as regards preservation. 


When iron is completely converted to massive oxide and no free 
metal remains, the specimen has been described as stable and no 
laboratory treatment is required for its preservation. When the rust 
is in a granular rather than a massive condition, however, there is 
always the possibihty of disintegration due to the crystallizations of 
salts, and in this case the specimen will have to be washed free from 
salts before its stabdity can be assured; the surface may have to 
be consohdated with nitrocellulose solution before washing as is 
done in the case of brittle pottery and stone (see p. 299). 

Rust should be preserved when its removal might be attended with 
the coUapse of the specimen. It should also be retained if its removal 
would be likely to result in serious disfigurement. A common ex- 
ample is the spotty surface sometimes found on a reduced iron ob- 
ject; the spots are due to compacted masses of ferric oxide occupying 
pits in the metal, and if these were dug out, the smooth surface would 
be made unsightly, and they are, therefore, left in situ. The risk that 
corrosion might break out again is extremely shght imder museum 
conditions, when salts have been removed as far as possible, and the 
object has been given a protective coating of wax or lacquer. 




s 2 

c . 


TJ g 



jj fco 


•a cu'H. 

^• 2-“ -3 
"^■u S 
a.S-2 = 
a u) p-S 

’Si-i« ° 

C «<> 2 

•S 3 • *3 
S'S.S S 

c« O 
.£ rt 

*j - -a 
c r= ** 

S -5 = 


2 i ^ a 

*- n3 « 

"3 "41 

*c a tj 

X Si “-g 

I S.s s 


3 £! 2 -2 

0 a •§ 8 


1 sjs-a 

a "O SP >, 

2 "O 2 2 

j3 a , 1 , 
° 2 
S’g J 

§ 2 « 

^ =§1 
« .s 

S -o -a 

S 2 

V I 

’T3 ;t 3 o 

H !5 rt 
"3 "a ^ 

'O M • 

o c .3 
^' o U-. h 
g.ti o ii 

'O >^ 

-213 3 
ja tiH 2 
-, *0 
U rt 1) ^ 

-a ^ j3 o 

"3 S ^ 

2 s 

3 -“ 2 ' ” 

2 s o S 

g :: Ml a 




»-• H c! 

- .S -S 
c -S ^ 

s.s g 


uh a-o 2 

o « -s &, 

_ u u 

^ j3 

S c) *5 

J2 .ts «1 

i-i rN <« 

« 5 *-• 

S-rf ° 2 

^ rt o” 
° >■> *0 

M -ua I>H • 

•S O «j ^ 

:2 g S'— ^ 

^ t; u^-S 

± o jc .a *5 'IS 
a Oh t-> G « O 

g “ u S a >- 

3^ 0-5^ o 

«j_i ^ 4i «s 

C Q ^ ^ <« 

n .2 

>s o ^ 

>s ? >-l M .2 S 


§ 2.2^ 5 ^ I 

W3 va a ^ ^ 
rt 0 H ClTO -ti 

2 u . S O ^ 

« <n O 
^ %M 4) j3 

•5 o g o-a — 

3 i>v— <3 -G t) 

>' 2"® ^ fi 2 

tj ^ ‘«w 0 B 

»> 9 >. c ^ a 

-o r.'s "-s “ 

•M 2 4) S C *fl 

-“ a -o u 2 *2 *2 
rt a (4-1 — 13 .a « 
33 S o a, « (L, Cm 

•- S — 
^ °s 

T3 cei. 

4> 4m 
* 2 ^ ^ 
;3 «« »G 
In CU 
« « 2 

n 3 








'33 w 5 

’ 5 b ^ 

4 ) "3 
t/i M 

4 J n 


.3 c 
> 5 

C 3 

0 3 

P 2 

s* ^ 

4 > ^ 

tj a 







o ^ 

3 'G *3 

is o y 

fl VJ ‘cu 

0 > 8 

<J ? 2 


Ts a o WJ 

u 43 J: s 
« ^ *3 
.§ S S-3 

1’"-® CM 

• S •-Um 



« t3 O 4> 

2 g *4 g 
to S > 
n tiS O 
(is n 511 u 

-C = !i 

m- 4 'a 

o ® _ a 
„ a.M4 a «1 
i.c °'^S 

. 2 ^ o-a 

«3 ^ i-a 
«.s 2 -§jSj 

^ 'c 'S 5 "c 
2 o 0 0 g 

t ) 

rs 1^13 
■G'2 S 
a g a 

o 14 a 

J 2 M. ^ 

•t: CM" 

a §-S 

— 8 -3 

"pM-, ^ 

« “"S 
■“ !>■ S 

•g.2 a 

a « S 
M n o 

a u “ 

gal 3, 
« o o 

rt ^ ^ 
.'X3 ^ 
G (J a 

O >» 

8 2 a 

« *13 c5 

rt G Q 

4 ) rt y 

^ cj 3 *■> 

o ^ S ^ 

T* ’R ^ 

^ >> W 

N *0 fG ,n 
a. S g? a ^ 

a. *5 2 -G g 

.. H b cLi a 

o « n 4J 

" “iS 

'C S -S" § 4I 

o\ ^ ^ 2 

^ On -a tJ y 

'-' F ro W ii 

^ C 4> 

.$• '3 "■ « a 

5 -S' S 2 u 

^ ^ i2 ^ S3 

‘ S TO vT 

1 ^ .3 ^ 

•> Gh 5 
^ 3 2 .D § 

^ -I ”h^ 

•G Fri 4 > t *4 2 
C/3 Sh ^ O « 
o O -W 4> ’3 
O O O to rt 

ffi « 2 

C 3 2 .3 
■ O O 3 4> > 
a ij O jOrg 
3 3 (rt 

c/3 (/} — * bI34M 

a s -c-g 2 

»G "G 4/ is 4 ) 

r r 

.® = j1.g 

3 fl *-» • ,*^ 

o o o ® ^ 

t S' 3 > 

J2 rt T® >.2 


w '♦• « G 2 

4) G 

o a 

(Plato, Crilias, 114 e) obtained from a naturally occurring mineral. 



Iron rust may have an evidential value. It is never safe to proceed 
to reduce a rusty object without making a careful examination to 
ensure that there are no features pecuhar to the rust that will be 
sacrificed in the process. Thus, a rusted shield grip may retain firag- 
ments of a leather belt, or the oxidized surface of a sword may pro- 
vide the only evidence of the structure of the scabbard or the hilt. 
Textile imprints in rust may be important, and even the off-prints 
from reeds or grasses with which the corroding object was in con- 
tact. Compacted rust is to be regarded, therefore, as a possible source 
of evidence that might otherwise be unobtainable, and laboratory 
treatment, on occasion, may have to be directed to the preservation 
of this evidence, rather than to the restoration of the object as a 
museum specimen. 


In the absence of a personal examination, it is never possible to give 
more than an indication of what treatment would be hkely to be the 
most desirable in any particular instance, but some further guidance 
may be obtained by studying case sheets. Brief practical notes are 
added relating to a dozen iron objects selected as being representative 
of different stages of minerahzation, in the hope that they may illus- 
trate the appUcation of methods already described, and at the same 
time be useful for reference; see Table VI, p. 282. 


Iron is unique among the metals of antiquity in that it may be 
decomposed to fragments of rust that are stable in themselves, and 
not subject to further corrosion. It is sometimes possible, therefore, 
by studying the fragments and fitting fractured surfaces together, to 
reconstruct an object, in part or whole, and by this means recover a 
specimen of archaeological importance from apparently worthless 

Two of the tabulated examples. No. 7 (the helmet) and No. 12 
(the axe), were received as a mass of fragments, some friable and en- 



crusted with sand, others very hard, and partly transformed into 
hmonite — a hydrated oxide of iron. In each case the laboratory task 
was to discover the shape of the object, to assemble the fragments and 
effect as far as possible a reconstruction. Such work is a variation of 
the famihar jigsaw pu2zle, but it must be carried out in three dimen- 
sions, instead of two, and since many of the pieces are missing, every 
scrap of evidence has to be marshalled before any progress is possible. 
Thus no preliminary form of cleaning can be undertaken, as this 
would destroy the natural variation in the colour and texture of the 
rust that helps to indicate which pieces come together. 

The first step is to set aside in groups all pieces bearing fragments 
of decoration, all pieces having distinctive grain markings or stains, 
and pieces that are of the same thickness and may belong together. 

Maryon^ describes an early stage in the reconstruction of the 
Sutton Hoo hehnet as follows: 

When sorting the fragments, I found that to facUitate the handling of 
each piece, and to preserve any dehcate edges, it was a good plan to pro- 
vide each fragment with a piece of stiff card, upon which it could rest, 
and by which it could be hfted. On each card the outline of the fragment 
was drawn, and notes made of details of any ornament or other feature 
of interest which had been observed. Search through, and study of, the 
recovered fragments, continued for a long while, but gradually some more 
general observations became possible. 

This method of dealing with the fragments on cards is invariably 

Attention is next devoted to fragments containing ornament, and 
when pieces are found that fit together they are stuck in position with 
Durofix. In such reconstruction work a box of sand is very useful 
for holding pieces at any required angle whilst the adhesive is harden- 
ing. In this manner and with the exercise of much patience forms 
begin to take shape. 

There comes a time when no further pieces can be made to fit. At 
this stage in the reconstruction of the Sutton Hoo hehnet, all 

' Maryon, H., Antiquity, 1947, 21, p. 137. 



discarded rust, clay, &c., was re-examined, and it was found that cer- 
tain shapes had been impressed on worthless material that had once 
been in contact with the helmet. Impressions of these shapes were taken 
in plaster of Paris, and the plaster models were found to be the equiva- 
lent of some of the pieces missing from the hehnet. A fresh impetus 
was thus given to the work — ^and when, eventually, certain puzzling 
fragments were recognized as being the remains of cheek pieces and 
neck guard, it was clear that a satisfactory reconstruction was within 
sight. The original material when brought to the laboratory con- 
sisted of ‘a gilded bronze nose and mouth-piece, two gilded bronze 
dragon heads, parts of what once had been a silver crest, and three or 
four hundred fragments of sand-encrusted rusty iron’. The latter 
might long have remained of httle significance had it not been for 
the development of the reconstruction technique that made possible 
not only the recovery of this helmet, but several other important 
objects from Sutton Hoo — ^notably the axe, the standard, the shield, 
and the drinking-horns. 


Metals that are uncorroded may be joined together by soldering, 
i.e. by alloying them with another metal of lower melting-point that 
can be made to flow through the joint and, on sohdification, act as a 
bond of luiion between them. Full details of the process are given in 
textbooks on metal-working,^ and it will suffice here to state some 
general principles that apply to soldering as a method of repairing 

There are two principal kinds of solder, hard solders which may 
contain many different metals and melt above 600° C., and soft 
solders that are composed of a mixture of lead and tin. It is only 
the soft solders that are of interest to us, and, in particular, the material 
known as Tinman’s solder. Tinman’s solder Grade K (B.S.S. 219) is 
of general utility and contains three parts of tin to two of lead. Its 
melting-point is so much lower than that of either tin or lead that it 

e.g. Maryon, H., op. cit. 



can be used for repairing either of these metals as well as all other 
metals of antiquity. 

The surfaces to be joined must be clean, i.e. free from oxidation 
and tarnish and any other impurity that will prevent the molten 
solder from coming into intimate contact with the metals with which 
it is to unite. But even the heat required in the process of melting soft 
solder would cause the metal to become oxidized unless it is pro- 
tected; it is necessary therefore in the process of soldering to use an 
oxide-solvent, called a flux. Fluxes are of two kinds: the first contain 
chlorides such as zinc chloride and ammonium chloride, while the 
second include those based on resins and fats. The former are very 
effective but difficult to remove completely afterwards, and for this 
reason are likely to give rise to subsequent corrosion; the latter are 
less efficient but do not cause corrosion. 

In repairing antiquities, the non-corrosive fluxes are to be pre- 
ferred, but we are then faced with the difficulty that while the 
old-fashioned resin fluxes are adequate for work on new metal (as in 
joining electrical cables, where absence of corrosion is an important 
factor), they are intractable where there is intercrystalline oxidation 
as in the case of ancient metal. A compromise is thus necessary, and it 
is to be found in the proprietary materials sold as ‘non-corrosive’ or 
‘activated resin’ fluxes. These still require the joints to be clean, but 
are easier to work than the simple resin fluxes. 


As it is important to have a working knowledge of the properties 
of fluxes used in the repair of metals, the following notes relating to 
some of the commonest that are commercially available may be of 
interest. Comments are based upon workshop experience rather than 
laboratory testing. 

Corrosive fluxes 

(i) Baker’s Soldering Fluid. This is recommended for general-purpose 
work, such as repairs to tin, brass, copper, or iron articles, and for 
joining and backing-up electrotypes. The joint requires thorough 



(2) Fryolux. This solder paste is very useful for timiing two surfaces that 

have to be closely fitted, such as a piece of brass tubing telescoped 
into another piece. Useful for ‘sweating’ screw heads into their 

(3) Multicore, ARAX. This flux is very good for general-purpose work 

and can be used with stainless steel. 

(4) Solderine. This does not have the same self-cleaning properties as 

I, but is satisfactory on surfaces that can be very thoroughly 
cleaned before apphcation. It is less corrosive than i, 2, and 3. 

(5) Coraline. This solder paste is similar in action to 4, but possibly 

rather less corrosive. 

Non-corrosi ve fluxes 

(6) Multicore Solder Paste, Cored Ersin Solder, and Liquid Ersin Flux. 

AU three have been found satisfactory on cleaned surfaces, the 
Cored Ersin Solder being ideal for electrical contacts. 

(7) Alcho-Re. This solder paste has been foimd to be satisfactory when 

appUed to surfaces made spotlessly clean beforehand. 

When ancient metal is found to be very difficult to solder, the 
best flux to use is probably a mixture composed of 70 parts of zinc 
chloride and 30 parts of ammonium chloride. This has the advantage 
of melting at 180° C., but it is very highly corrosive. When it is used 
in making a joint, the object should be washed immediately after- 
wards in water acidified with a few drops of hydrochloric acid in 
order to decompose any insoluble zinc oxychloride that may remain 
on the joint, and it should be thoroughly washed finally in water 
made faintly alkaline with a httle wasliing soda. 

Soldering is required most frequently in repairing brittle metals, 
e.g. speculum or ancient silver, or in mending fine work such as that 
on brooches, handles, &c., where leverage would tend to rupture joints 
made by adhesives. Soldering is often used as a temporary fixing in 
‘tacking’ cracked metal together as is done in the reconstruction of 
crushed metal objects. When the original shape has been re- 
established, a permanent soldered joint can be made to take the place 
of the temporary fixings which are easily removed. (Cf treatment 
of the Emesa silver helmet, p. 226.) 




In exceptional cases it may be necessary to apply a modem metal 
patch to an antiquity. This should bear an unobtrusive but distinctive 
mark to indicate that it is modem; a small depression, for example, 
such as can be made by using a spring-loaded centre punch. 

Soft solder can be given a coating of copper by degreasing the 
solder with trichlorethylene and then rabbing it with a moistened 
crystal of copper sulphate. The copper surface thus obtained is less 
obtrusive than the white metal. Soft solder that has been appHed to 
silver can be concealed by degreasing and then silver-plating. This 
is done by ‘ragging’, i.e. plating with a glass rod anode around which 
is wrapped silver wire in contact with a small piece of soft rag mois- 
tened with a cyanide electrolyte.^ The solder is mbbed with the rag 
whilst a small current is passing (at about 3 volts) and it gradually 
becomes covered with a thin coating of silver. 

^ For the composition of solutions and methods of operation see The Canning Handbook 
on Electro-plating (W. Canning & Co. Ltd., Birmingham, i8). 

B 6157 






Several objects 
corroded together 





^ 3 


^ o 

*0 O C; 
S Jj o 

J ^"3 

,• * to 

•5 s cr 2 

c ' “ 


s m 




"o ^ 



£- o 

"D .s 2 

s «« 

« 2 4> 

. ^ u is 

s € o £P” 

rp^ u n rp 

« « 3 *£ o 


. a> 

6 J 5 > 

5 ® 
’E. S 
a- 1> 




rt tL. 2 


c ^ ^ o 

4 j — 3 C 

o = = c 

S s 

2 iJ c o 

« O « 
Oh 0^ i-J Q 


« C y -- 

:2t 3 tTN J 

u 'ij c ^ ^vo 

» rt -Q 

i.y ■>? -- 

I ij o c 

« S 2 « •« 

5 « c O C o ; 
> K w o > j 

'o O ^ ^ t> g t: 
i^sSSl O 

P5 CsJO- 

5 ^ tC > 

3 • C > 



2 S- 
cu « .5 

• - .^"6 •« 
w rt 

o- w 

•S £ 

<3 c 








Since early in the present century the preservation of stone has 
engaged the attention of scientists the world over. The phenomena 
of decay are commonly to be seen on pubhc buildings and outdoor 
monuments: weathering, which weakens the stone though it may 
sometimes enhance its appearance; staining, which may result in 
permanent disfigurement; and the crystaUization of salts, which in 
some cases may cause powdering of the surface, cracking, and even 
complete disintegration. Deterioration may be traced to bad technique 
such as carelessness, or lack of knowledge or experience in the hand- 
ling of stone by architects or builders, but it is more often due to 
natural causes. In this case, the most that can be done is to try to 
find a way of mitigating the damage, by washing, or periodic steam 
cleaning, or, it may be, by some form of impregnation. 

While some reference will be made to the treatment of outdoor 
stone monuments, it is principally the preservation of stone objects 
indoors that will be dealt with here. This is the simpler task, and 
although it introduces problems, it may fairly be claimed that most 
of them are capable of solution. 


The earth’s crust is composed of minerals associated together to 
form rocks, and, while the rocks have been used as such for build- 
ing and sculpture, certain of the minerals have attracted the artist- 
craftsman from earhest times because of their special quahties for fine 
work, e.g. hardness, texture, and colour. In the Far East and in the 
Americas, minerals such as jade, rock crystal, and malachite were 
used for fine carvings; in the Middle East, lapis lazuh, turquoise, and 
camehan were employed extensively for inlaying furniture and for 



the decoration of metal objects; and alabaster provided the Egyptian 
craftsman with an attractive material from which to fashion the 
lamps and unguent pots that were an essential furnishing both of the 
temple and of the tomb. 

Minerals, as a class, may be regarded as inert, and objects fashioned 
from them are unlikely to need attention other than periodic clean- 
ing, which can be done in some cases by washing with warm soapy 
water, or by brushing with spirit. 

There is, however, one exception, and this is the mineral called 
marcasite which is sometimes used for inlays. Marcasite is a sulphide 
of iron, being a variety of iron pyrites, and it has a similar brassy 
appearance. It is very hable to decomposition which takes place as a 
result of oxidation in the presence of moisture, the sulphide being 
converted to ferrous sulphate which forms a white feathery incrusta- 
tion on the surface, sulphuric acid being produced at the same time. 
Treatment may take the form of washing, or, in bad cases, soaking in 
dilute anunonia prior to washing; and since the acid is, of course, the 
major factor in causing decomposition, washing is continued until 
acid can no longer be detected on testing the wash water with litmus. 

When fossils contain decomposing marcasite and calcium carbon- 
ate they are often beyond recovery, because the acid resulting from 
the decomposition of the marcasite reacts with the calcium carbonate 
causing disintegration. Such fossils are said to be ‘pyritized’, and the 
best method of treating them is to expose the fossd to the fumes of 
strong ammonia in a closed vessel. This neutralizes any free acid, but 
the treatment may take many months and even then can never be 
guaranteed to be completely successful as it may not be possible for 
the ammonia fumes to reach all of the acid. After washing in water, 
drying is conducted quickly through baths of alcohol and ether, and 
the fossil is then impregnated with a lacquer to exclude air and 
moisture. For this purpose a sdicone lacquer^ is very suitable, as it is 
a water repellent; good results have also been obtained by using poly- 
vinyl acetate in a mixed solvent of nine volumes of toluene to one 
volume of acetone. 

* Midland Silicones Ltd., 19 Upper Brook Street, London, W. i. 




Of the igneous rocks, granite and basalt are most widely distri- 
buted. These are generally hard, non-porous, and very stable under 
a wide variety of conditions. Granite contains over 66 per cent, of 
sihca, and is an acidic rock, whereas basalt is a basic rock containing 
less than 52 per cent, of sihca. Between these two extremes there are 
many well-defined rocks of an intermediate sihca content. Many 
have been used for monuments in antiquity and have survived 
weathering extremely well. 

A problem is presented, however, when such stones are brought 
away fi'om their natural surroundings and exposed in the open to 
weathering of a nature that is foreign to them. Cleopatra’s Needle, 
a granite obehsk from Hehopohs, now on the Embankment in Lon- 
don, was observed to have suffered a marked deterioration in its 
condition on being taken from the dry atmosphere of Egypt to the 
humid atmosphere of England. A hke deterioration was observed in 
a similar granite monoHth that was taken from Egypt to New York, 
and it was found necessary in each case to treat the stone so as to 
prevent access of moisture. 

A study of the London monolith made in 1952^ showed it to be a 
typical granite composed of crystals of quartz, felspars, hornblende, 
and some white mica. While the stone was originally of a pink 
colour, it had become badly stained with a black film . This proved 
to be mostly carbonaceous in nature, but it contained an inorganic 
residue which consisted mainly of iron oxide and sihca. Light colomed 
patches marked places where flakes of the stone had become detached 
by the action of frost which was regarded as the prime cause of the 
deterioration. Conservation involved cleaning, drying, and water- 
proofing the stone. Cleaning proved to be difficult; the roughened 
surface of the granite necessitated the use of scrubbing brushes and 
even wire brushes, and it was found that the appHcation of an organic 
solvent was essential to soften the black film before it could be re- 
moved. The solvent was composed of nine volumes of carbon tetra- 

’ Burgess, S. G.. and Schaffer, R. J., Chemistry and Industry, 1952, pp. 1026-9. 



chloride and one volume of benzene emulsified by the addition of 
a surface active agent, e.g. i per cent, of Lissapol N. When the 
cleaning was completed and the natural colour restored, the stone 
was hosed down and allowed to dry; it was then given two appHca- 
tions of a lo per cent, solution of paraffin wax (M.P. 50° C.) dis- 
solved in white spirit. After the solvent had evaporated, the stone 
was warmed gently, using a blow-lamp, with the object of getting 
the wax into the cracks and porous sections, and thus excluding mois- 
ture. This worked well, and the stone, inasmuch as it has been water- 
proofed, is now protected ftom the action of frost which was the 
main object of treatment. The staining by soot, however, cannot be 
prevented, especially with a surface as rough as this Egyptian granite, 
and within three years the stone had become as black as ever. 

Basalts are more homogeneous than granites, and they seem to be 
less hable to deterioration. Certain monuments of lava from British 
Honduras were observed to be very crumbly on the surface after 
having been exposed in the portico of the British Museum for a few 
years and had to be brought indoors. These stones are porous and 
therefore prone to damage by frost, but how far the industrial atmo- 
sphere may have contributed to their deterioration is hard to assess. 

Such incidents serve to illustrate the general principle that when an 
objea is exposed out of doors it is hable to suffer deterioration from 
causes that are beyond control, and although the object may never 
have been intended for a museum, it may be necessary to keep it 
indoors so that it will be protected from dirt and the rigours of a 
climate that wiU hasten its destruction. 


The problem of preserving sedimentary rocks is essentially one of 
overcoming surface frailty. The particles of grit of which they are 
composed are cemented together in a matrix which, if disrupted, leaves 
the particles free to fall away in powder. On the other hand, some 
stones develop a natural skin or patina which may afford a certain 
degree of protection. If this skin is damaged, it exposes a powdery 
under-surface, and experiments made in an effort to repair such a skin. 


or to apply an artificial protective skin, have proved to be unsuccessful 
as a means of conservation. Indeed, studies in stone preservation have 
brought out the fact that, as often as not, the apphcation of im- 
pervious surface films to stone (as distinct from deep impregnation) 
seems to intensify deterioration instead of affording protection. 

The deterioration of the surface of sandstones and hmestones which 
are comparatively porous is caused mainly by the crystaUization of 
soluble salts. When these are present in the stone, or, indeed, in any 
porous substance, they will work their way towards that part of the 
surface where there is greatest evaporation; they will become con- 
centrated there, sometimes as a hard deposit but often in the form of 
filamentous crystals which appear as if extruded from the pores 
(Pis. 50 -A. and b). The growth of the crystals and the accompanying 
pressure may impose so much strain on the stone that the surface 
is disrupted, ornament is defaced, and inscriptions gradually become 
illegible. This can be prevented if the salts are extracted either by 
washing or by using a process involving the use of moist paper pulp. 

I . Removal of soluble salts by washing 

In specimens where the salt has not yet started to crystallize on the 
surface, the stone should be immersed in running water and finally 
in changes of distilled water to remove the salt before it can do any 
damage. When, however, crystallization has occurred on the surface, 
the salts should be removed as far as possible by dusting them off 
gently with a small soft brush prior to washing. When once the stone 
is wet, it should not he allowed to dry until it has been ascertained by 
means of the silver nitrate test that all but traces of chloride have 
been eliminated (see p. 199). 

Unfortunately, it is not always possible to treat specimens in this 
way. Statuary may be large and already weakened by salt action, and 
in the case of heavy stone the element of risk involved in moving it 
may not be justified. In the case of outdoor statuary it may be pos- 
sible to get rid of salts and dirt by using a hose, but for stones indoors 
that are bulky, the most convenient way of removing soluble salts is 
to employ the paper pulp method. 



2. Removal of soluble salts by the paper pulp method 

Paper pulp is available commercially, but it can easily be prepared 
by boiling soft paper with distilled water and beating it imtil com- 
pletely disintegrated and the cellulose fibres mat together forming 
a pulp. When cold, the paper pulp is thrown against the stone so that 
it adheres, and a kind of papier mache garment is gradually built up 
until the object is concealed completely (Pi. 51). The water in the 
pulp will he absorbed by the stone and the pulp will shrink shghtly 
as it dries so that, eventually, the stone will be covered with a porous 
layer of material about j to | in. thick. The water absorbed by 
the stone will dissolve the soluble salts, and, at first, carry them 
farther in; but, because salts in solution tend to move towards a sur- 
face where evaporation is taking place, they will in time change 
direction and eventually leave the stone, forming an incrustation in 
the pulp. The pulp is left in position for about three weeks, when it 
may easily be removed; it is then replaced with fresh pulp and if the 
stone is very salty a third appHcation may be necessary. 

In cases where the surface of the stone has been loosened by in- 
crustations of crystals, or where there is painted decoration, as in 
Egyptian tomb paintings, consoHdation is necessary before applying 
the pulp. This is done by painting the stone freely with a 2 per cent, 
solution of celluloid in a solvent consisting of equal volumes of amyl 
acetate and acetone.^ By this means the powdery grains are held 
together and the paint fixed, and when quite dry the stone is ready 
for the pulp treatment. The film of nitro-ceUulose does not prevent 
the extraction of the salts, but as the diffusion takes place more slowly 
it may be necessary to continue the treatment for several months, re- 
newing the pulp coating from time to time during this period. If the 
painted decoration appears to be shiny after treatment, the gloss can 
be removed by dabbing the areas in question with cotton-wool 
moistened with a Uttle acetone, but this should be done with care as 
it is hable to dissolve too much of the nitrocellulose and possibly 

’ While the paint may become glossy by such treatment, stones that contain or have 
contained much salt are generally very porous and can absorb large quantities of celluloid 
solution without change in appearance. 


loosen the consolidated surface. It is usual for the texture and appear- 
ance of the painted surface to be improved after treatment by the 
paper pulp method and sometimes they may even be enhanced by a 
final coating of a 2 per cent, solution of celluloid. It should be noted 
that although the paper pulp method can be extremely effective, it 
does not remove all the soluble salt from an incrusted stone; a 
certain residue is bound to remain, but if the work has been carried 
out properly this will have no material effect upon the stability of 
the specimen when it is preserved under museum conditions. 

As an example of what may be achieved through the successful 
apphcation of this treatment we may instance the case of an Egyptian 
Mastabah tomb of limestone which was firail and encrusted with 
salts. As a preliminary step it was consohdated with a solution of 
celluloid. This was followed by three apphcations of paper pulp 
spread over a period of four months. By this time the incrustation 
had disappeared and what remained of the natural surface of the 
Hmestone was exposed, bringing to hght deHcate carvings in has 
rehef and painted decoration which had long been completely 
concealed from view. 

3. Removal of insoluble salts 

Sometimes the prevailing incrustation is insoluble in water so that 
washing or treatment of the stone by the paper pulp method are of 
no avail and alternative methods must then be employed. These will 
depend on the chemical nature of the incrustation. If the salt is car- 
bonate, it can be decomposed by any acid, but a warning is necessary 
since even sandstones may contain calcite (calcium carbonate) as an 
essential part of their structure; before using acid, therefore, it must 
be estabhshed that the rock is neither a limestone nor a sandstone 
containing calcium carbonate. The rock, when tested, should not 
itself effervesce with acid. Even when all seems well, acids should 
only be apphed locally and sparingly, not with the object of dissol- 
ving salts wholesale, but rather with the object of loosening the 
crystals so that they may be dislodged mechanically. For this 



purpose, dilute hydrochloric acid {5 per cent, or even weaker) may 
be used, followed by thorough washing with water. 

When the incrustation is selenitic, i.e. when it consists of gypsum 
or calcium sulphate that has been laid down in glassy crystals, it may 
be very disfiguring. Such incrustations have formed very slowly and 
are probably harder in the mass than the surface of the stone which 
wiU certainly have been weakened during their growth. It is imwise, 
therefore, to try any chipping or grinding process until preliminary 
steps have been taken to try to soften the incrustation. Unfortunately, 
no solvents are available that will dissolve gypsum readily. If feasible, 
the best treatment is to keep the stone immersed in lukewarm water 
for a long time, changing the water every 24 hours; this has been 
found to be more effective than immersion in either cold or hot 
water. Solutions of sodium thiosulphate (10 per cent.), or ammo- 
nium carbonate (10 per cent.), may be apphed locally, but these 
chemicals act slowly and they must be washed out very thoroughly 
afterwards or the cure may be worse than the disease. 

A novel method of treatment that at times has proved to be very 
effective depends on the fact that selenite is a hydrated mineral, and 
that it goes to powder when it is dehydrated by heating. An incrus- 
tation may sometimes be heated with an electric soldering iron, and 
if the temperature is strictly under control it is possible to disinte- 
grate the gypsum so that it can be brushed away without damage to 
the stone. The method is obviously not without danger, but in cases 
where a fine object is suffering fiom a gross disfigurement of selenite 
which would otherwise be difficult to remove, it may be worthy 
of consideration. Needless to say, the stone must be quite dry when 
heat is applied, and it should be borne in mind that overheating a 
limestone converts it into quicklime. 

4. Consolidation 

After salts have been removed from stone, it is often necessary to 
strengthen the surface. In the British Museum Laboratory hardening 
tests were carried out upon a number of Coptic stele of very porous 
hmestone that had been desalted by the pulp method, using such 



consolidating liquids as wliite beeswax in turpentine, white shellac 
in alcohol, and solutions of cellulose nitrate and polyvinyl acetate. 
It was found that, under museum conditions, success was related 
directly to the degree of penetration of the consohdating agent. After 
a lapse of twenty years a further series of observations was made on 
the same stele after they had been exposed to damp in an unheated 
war-damaged building. This confirmed the earher observations that 
the most satisfactory consohdating agents are those which can be 
made to penetrate well into the material and not merely remain 
on the surface, especially if the stones are hable to be exposed to 
damp and varying temperature conditions. 

It is clear from the tests that it is not possible to preserve a powdery 
stone surface by merely giving it a coat of varnish. Porous stone 
contains air that will expand and contract with variations in tempera- 
ture, even under museum conditions, and no thin surface film could 
long withstand the consequent movement without fracture. This is 
the reason why the most successful results have been achieved either 
by securing deep penetration of the consohdating agent, or by em- 
ploying a form of surface treatment that leaves the stone free to 
‘breathe’. These treatments will now be described in detail. 

(i) Impregnation with wax. If properly carried out, wax treatment 
has been shown to stand the test of time satisfactorily. Where possible 
the stone should be immersed in a tank of molten wax — a method of 
treatment that has already been described in the case of wooden 
objects (see p. 127). Special plant is required for such treatment, how- 
ever, and the size and weight of the stone might be such as to make 
this method impracticable. When wax has to be appHed to the sur- 
face of a large object that cannot be immersed, the only way to ensure 
penetration into the pores of the stone is to heat the stone for a time 
before applying the wax. The stone must, of course, be thoroughly 
dry before it is heated. 

A convenient method of heating is to use an electric radiator placed 
about three feet from the stone. By this means, a wide area can be 
dealt with at once. When the stone is hot, the radiator is switched off 
and white beeswax appHed in the form of a thin salve having the con- 



sistency of vaseline. This salve is prepared by stirring molten beeswax 
at about 85° C. into petroleum ether (B.P. 80-100° C.), care being 
taken not to do so in the presence of naked Hghts. When the salve is 
apphed to the warm stone, the wax is absorbed into the pores and the 
inflammable solvent evaporates. After all the solvent has evaporated 
heating may be repeated, and further coats apphed in the same man- 
ner as long as the wax is being absorbed. 

Perhaps it should be mentioned, in the interests of aesthetics, that 
waxing invariably causes a certain dullness and lowering of tone 
particularly noticeable on hght stones, and also that waxed stone is 
easily stained by dirt. This latter problem is not so serious when 
microcrystalline waxes are employed (see Appendix XII). 

The waxing method may be used also for statuarv^ that is kept out 
of doors, but it is absolutely essential that the stones should be quite 
dry before they are waxed and that the wax should penetrate well. 
Where stones are set in a wall, wax should never be used as a con- 
sohdating agent as there is always the possibflity of water getting into 
the foundation or behind the wall and causing future trouble. An 
examination of many frescoes on the walls of pubhc buildings and 
churches has shown, without doubt, that treatment with wax has 
only been successful where the fabric of the building has been kept 
continuously dr\\ Where water has had access to the walls by the 
chance blocking of a gutter, or perhaps through inadequate drainage 
on the outside, the moisture cannot escape through the waterproofed 
surface; it is held back by the wax barrier where it may cause de- 
terioration at the interface, with the result that large flakes become 
detached, or, in some cases, even the foundation layers of the fresco. 

(2) Impregnation with lacquers. In attempting to impregnate a stone 
with film-forming solutions (lacquers), one has to face the difficulty 
that any solvent that goes into the stone must eventually evaporate 
and come out again, at the same time bringing much of the sohd 
content of the lacquer with it to the surface, or the immediate sub- 
surface. This is easily demonkrated by cutting sections of a white 
porous stone that has been impregnated with a coloured lacquer: it 
will be seen that the colour has penetrated well into the stone, but 


that the sohds remaining after the solvent has evaporated are distri- 
buted near the surface. Lacquering methods might, therefore, seem 
to be of limited use, but, where apphcahle, as when stones are kept 
indoors, they can be regarded as satisfactory. It must be emphasized, 
however, that this method of treatment affords no protection to 
stones that are exposed in the open to the undermining action of 
rain and frost. 

Lacquer solutions must be appHed with a brush in a series of coats, 
the first very dilute, and the succeeding ones progressively more 
concentrated, allowing time for each coat to dry before applying the 
next. As has already been mentioned, white shellac gives good results. 
This is obtainable as French polishers’ ‘white polish’, and is diluted 
with methylated spirit for use on stone. Other impregnating media 
that have proved satisfactory for indoor work are polyvinyl acetate 
in toluene/acetone solvent, and Bedacryl 122 X* in xylene/toluene 

(3) Vacuum impregnation. For small stone objects that are in a 
powdery condition, the best treatment is to impregnate them in 
vacuo with one or other of the above media well diluted with sol- 
vent. This is carried out in a vessel that can be made air-tight and is 
at the same time strong enough to withstand the strain of being 
evacuated. For small work a thick glass desiccator^ serves well (see 

Fig. 5, p. 152). 

The stone is immersed in the impregnating solution, air is sucked 
from the vessel, and bubbles are seen rising from the stone through 
the solution as the pressure falls. After most of the air has been 
removed and the bubbling has subsided, air may then be readmitted 
to the vessel. The pressure of the incoming air will force the solution 
into the pores of the stone. It is most important that the air should 
be admitted gendy, as a sudden inrush might cause the stone to 
rupture. After impregnation, the object is allowed to drain above the 
solution in the atmosphere of the solvent to reduce the rate of drying. 
Hasty drying is Hable to bring too much of the lacquer to the surface, 

“ Imperial Chemical Industries Ltd. 

* See also Burgess, M. E., Museums Journal, 1954, 54, p. 125, 


Surviving elements dispLivctl in a Perspex mount 



leaving the stone with a shiny appearance. The apparatus has been 
described in greater detail under the impregnation of wooden objects 
(see p. 130). 

(4) Silicon ester spray. For slabs of sandstone and sihceous limestone 
of large dimensions which are kept indoors, a most successful 
strengthening agent is silicon ester, but it is essential that it be appHed 
strictly in accordance with the manufacturer’s directions. 

Silicon ester^ is available in the form of a dilute solution in alcohol, 
and it is necessary to apply it with a spray or an atomizer as it is 
not effective when merely brushed on the stone. 

The atomizer should be held at some distance from the stone, so 
that the liquid reaches the surface only as a fine mist. Further apph- 
cations are made not oftener than once a week, thus allowing time 
between each operation for the ester to decompose and coat the 
grains of stone with sihca. After about three appHcations the surface 
may be tested by rubbing tightly with the frngers, and if the dust has 
been fixed, this will indicate that the process is tikely to prove success- 
ful. Eventually, after about three to seven applications, a trace of per- 
manent milkiness will be observed — a sign that treatment should be 
discontinued, as the surface of the stone is now well consolidated 
without the pores being clogged. Sandstones so treated have been 
found to remain in good condition in the museum for many years. 

The method is not so effective with fine limestones, though in some 
cases it seems to harden them. Disappointing results were found in 
the case of lavas and it would seem that the type of stone that responds 
best to this spray treatment is one having a gritty surface, the silicon 
ester reinforcing the decomposed matrix and thus consolidating the 
gritty constituent of the stone. 

It should be added that there is a tendency for silicon ester to clog 
the jet of the spray, and once this happens it is difficult to clear it for 
use again. The spray should therefore be dismantled between opera- 
tions and the jet washed with alcohol and kept, preferably, in a bottle 
of spirit. 

^ A suitable brand of silicon ester for the preservation of siliceous stones in museums is 
that known as Silicaseal No. ia, obtainable from Silicaseal Ltd., Newcastle upon Tyne. 

B 6157 




Silicon ester has a comparatively short shelf life, and it should be 
used within a few weeks of purchase, as even dilute solutions become 
viscous and unfit for use after a lapse of a few months. 


Both igneous and sedimentary rocks may be changed in form, 
depending on their composition and history, granitic rocks becoming 
gneisses, sandstones quartzites, and limestones marbles. It is the 
marbles that are of most interest in museum work as they have been 
used to such a large extent for sculpture as well as for building 

Marble is a metamorphic rock arising from the action of heat or 
pressure, or both, on limestone. By this action the original character 
of the limestone is lost, and it becomes an aggregate of calcite crystals 
which may be pure white, coloured, veined, or black, according to 
the impurities present. As a result of this transformation the pore- 
space in the stone is reduced and is smaller than that of hmestone, and 
marble is thus capable of taking a higher pohsh than limestone; but 
the pores are not so small as to prevent the marble from being easily 
stained, and special care is necessary to prevent the accumulation of 
dust in the hollows of white marble sculpture, as ingrained stains may 
be very difficult, or even impossible, to remove. 

Dusting and washing. For the preservation of marble in the museum, 
periodic dusting is essential, and this should be done with a large 
feather whisk or soft brush. Cloths should never be used as they tend 
to rub the dust into the stone. 

W ashing may be necessary at times, and for this purpose it is im- 
portant to use water that is free from iron, and preferably distilled. 
A Httle good-quahty soap may be used, just enough to cause a shght 
frothing when apphed with a soft brush. The solution used in the 
British Museum for the periodic washing of marbles consists of: 

Soft soap (B.P.) . . .10 grams. 

Distilled water . . .100 ml. 

Ammonia (0.88) . . .1 ml. 



This is made up as required, kept in a glass bottle, and used from a 
glass or pottery vessel. The mixture must not be allowed to come in 
contact with iron, and should not be used from an iron pail, even if 
galvanized, as this might lead to eventual staining. After the marble 
has been dusted with a feather whisk, washing operations are com- 
menced at the top of the statuary. The cleansing solution is worked 
with a soft brush into a froth on a small area at a time, taking care 
to prevent dirty water from lying in the hoUows or running down 
in rivulets over areas of unwashed marble. When clean, this area 
should be dried with a soft towel before proceeding farther. 
Only when all the marble has been cleaned and mopped dry is the 
stone washed freely with fresh water to remove the last traces of soap, 
which would soon collect dust again if allowed to remain. 

For marble that has been neglected and allowed to accumulate 
dirt for years, it may be necessary to use a stronger detergent, such 
as the froth from Lissapol, Teepol, or one of the proprietary cleansers 
made from such materials. A convenient quantity is 2 oz. of 
detergent in one gallon of water, and this concentration should 
not be exceeded. Such a solution is not for regular use and should 
only be apphed in exceptional circumstances, taking the same pre- 
cautions as before to prevent rivulets of dirty water from nmning 
down the marble. 

Removal of stains. The common staining ingredients of dust are soot 
and traces of iron, and these give rise to grey and to rust-coloured 
marks respectively. White marbles easily pick up colour from con- 
tact with packing materials such as paper and straw if these are 
damp, and even damp clean dust-sheets have been observed to cause 
staining. In one such case the source of the trouble was a slight 
mildewing of the textile which caused the appearance on the marble 
of a brown line following the course of the retaining string. The sus- 
ceptibility of marble, and particularly of white statuary marble, to 
such forms of staining makes it essential to take special precautions in 
packing. Where there is a possibility of exposure to damp, white 
marble should not be allowed to come in contact with any organic 
material liable to fungoid attack. This fact was brought home by the 


discovery of intensive polychrome staining all over a fine white 
marble bust of Voltaire by Houdin. The marble had been wrapped 
in a white silk eiderdown, boxed up, and stored in a garage where 
conditions were at one time so damp that the wood of the box had 
rotted. When the box was opened the packing was found to be con- 
cealed in a mass of mould growths, and the marble was stained 
brown, red, and green at all points of contact. Much of the staining 
was removed by the use of an aqueous 2 per cent, solution of chlora- 
mine-T, but certain areas of dark staining resisted all efforts to bleach 

Another type of staining by contact with organic material was that 
suffered by the Michelangelo tondo in the Royal Academy of 
Arts. In this case the trouble was caused by felt padding pressing 
against the white marble, and this resulted in the appearance of a rust- 
coloured smudge which was very disfiguring. Fortunately, most of 
the stain responded to washing and treatment with chloramine-T, 
but rather than take the risk of using a stronger reagent, the residual 
yellowing was, for aesthetic reasons, chalked over to conceal the 
blemish. In the course of time the yellowing became decidedly less 
intense. The gradual disappearance of the stain was not due in any 
way to the presence of the chalk but no doubt to the volatihty of 
the residual staining material. 

It is not an easy matter to restore the pristine whiteness of Carrara 
marble that has been stained. Although it may be possible to remove 
coloured matter from the surface, in most cases the stain will have 
penetrated into the stone where it can no longer be reached by sol- 
vents, and may remain as a permanent disfigurement. The problem 
of removing stains from marble is further complicated by the fact 
that all acids disintegrate marble, and for this reason the use of specific 
acidic solvents for removing stains is out of the question. The com- 
mon neutral organic solvents are safe to use, and slightly alkaline 
solvents are permissible, but when the latter are used they should not 
be left for long in contact with the marble and should be washed off 
very thoroughly afterwards as there is a possibfiity of their causing 
yellow stains by contact with colourless impurities present in the stone. 



When the marble is sound and the staining is general, thorough 
washing, and brushing in the wet condition, may be undertaken, but 
when the stain is localized and highly coloured, it is usually best to 
determine the nature of the stain and employ selective solvents with- 
out prior washing. Thus, a freshly prepared 2 per cent, aqueous 
solution of chloramine-T will remove stains caused by red ink and 
also ink stains of the blue-black type, though in the latter case a 
yellowish residue is likely to remain. This residual coloration may 
respond to treatment with hydrogen peroxide (20 vols.) to which 
a drop of ammonia has been added. After the stain is discharged the 
marble should be thoroughly washed. 

A common type of staining is that due to oil paint. A thick blob 
of hard oil paint should be carefully scraped off without damaging 
the stone, and any residue treated with a suitable solvent. Pyridine 
and morpholine are useful in dealing with bad oil stains, or staining 
of a bituminous character, as is also a mixture of equal volumes of 
benzene, ammonia (o- 88 ), and methylated spirit. These hquids are 
apphed locally with a stencil brush, and, after mopping with cotton- 
wool, the marble is washed dovm thoroughly with water. It should 
be noted, however, that the oily ingredient of the paint will certainly 
have been absorbed to some extent by the marble, and, even if it is 
possible to eradicate all of the surface pigment, the chances of extract- 
ing all the oil from the pores of the stone are very remote indeed, and 
a grey or brownish residue is almost certain to remain. This is not 
a serious matter if the area affected is in an inconspicuous part, but 
if the stain is a bad disfigurement it is much better to cover it over 
with chalk rather than to torture the surface of the stone with 
chemicals in the remote hope of improving matters. 

A method that is sometimes of value in removing surface stains is 
to use a stripping film according to the procedure developed for re- 
covering fossil imprints from coal. This consists in applying to the 
stain a viscous solution of nitrocellulose so compounded that, on 
drying, it leaves behind a vet)' elastic film; when this is peeled off 
it may take the stain with it. The method has given good results 
in the case of stained marble but could be apphed equally well to 



any other fine-grained stone. The following solvent mixtures are 
suggested by Duerden;* 

(1) Rapid drying: Alcohol 

Castor oil 

(2) Slow drying: Acetone 

Amyl acetate 

I vol. 

1 vol. 

5 per cent, by vol. 

2 vols. 

1 vol. 

2 per cent, by vol. 

A sufficient quantity of nitrocellulose (celluloid cuttings) must be 
added in each case to give a viscous solution. 

A common disfigurement is the formation of organic deposits 
caused by the growth of lichens, algae, &c., on the surface of 
marble exposed in the open. These deposits may be softened and re- 
moved by treatment with a Httle dilute ammonia. In cases where the 
algal growths are extensive, they may be arrested by spraying with 
formalin, and this treatment facihtates their subsequent removal. 

When the surface of a finely pohshed marble is blemished locally 
by an insoluble deposit the only cure may be to wet the marble and 
remove the deposit by rubbing with a small chisel-shaped shp of 
water-of-Ayr stone or snake-stone. ‘Snaking’, when apphed hghtly, 
does not impair the pohsh of the marble. 

Decomposition and consolidation. Marble has been described as an 
aggregate of crystals of calcite, and these crystals have a different 
coefficient of thermal expansion in two directions. The result is that 
when marble is heated it tends to become distorted. This distortion 
is detectable even below 100° C., but when the temperature is as 
high as 400° C. warping may be considerable. Such deformity is 
sometimes noticed on old mantelpieces, and, as it is a case of the 
stone having taken on a permanent set, there is no way of recover- 
ing the original shape. 

The second type of deterioration resulting from heat causing 
irregular expansion is the granulated condition sometimes observed 
on projecting portions of statuary. In this case, expansion at different 

' Duerden, H., Ann. Bot. 193 1, 45, p. 376. 



rates in different directions causes the crystals to lose their power of 
cohesion and it is possible to release granules pf the marble by merely 
rubbing such surfaces with the finger. The affected portions are a 
staring white, as they scatter the hght more than the surrounding 
stone. Examples have been found in the Parthenon marbles, and in 
marbles firom many other sources. In one case, an extended limb of a 
large Canova group was worn down to a shapeless mass of loosely 
adherent crystals having the appearance of granulated sugar. 

It is possible to do something to consohdate museum marbles in 
this condition, but the same difficulty arises as in the consohdation of 
sandstones and limestones, namely that impregnating agents requir- 
ing the use of solvents that evaporate cannot be persuaded to fill the 
pores of the stone because the solvents, having gone in, must find 
their way out again. For some purposes the most satisfactory treat- 
ment may be to impregnate the stone with wax, even though this may 
seem dlogical, as it involves applying heat to the stone. The aim, 
however, is to get the wax to penetrate well into the granular marble, 
and this can be achieved to a depth of an inch or more, if the stone is 
warmed by an electric radiator. The radiator should not be placed 
too near the stone, and sufficient time should be given for the stone 
to warm through gradually. After removing the radiator, treat the 
stone with a soft buttery salve prepared by stirring paraffin wax of 
M.P. 46° C. into petroleum ether (B.P. 40-60° C.). It will penetrate 
the granulated marble easdy, replacing the air and filling the pores so 
that the marble now has a more or less umform appearance and the 
enfeebled stone is consoUdated. As petroleum ether is highly inflam- 
mable, care should be taken to see that there are no naked hghts in 
the room during operations. 

Another method of dealing with granular marble in the museum 
is to impregnate the porous areas with lime water, giving three coats, 
and allowing a few days’ interval between each for the marble to 
dry. This is followed, if necessary, by one or two coats of a 10 per 
cent, solution of ‘soluble casein’. By this means the granules are 
consoUdated by a film that consists essentially of calcium caseinate 
and caldiun carbonate. In cases where the original surface of the 



marble remains to a large extent unimpaired, and there are ordy 
isolated granulated patches, there may be justification for stipphng 
the staring white of the granulated areas with a httle pastel colour to 
match the surrounding unaffected marble. This masks deformities 
and assists in recovering the aesthetic appeal of the work. (For repairs 
to broken marble see p. 3 17.) 

It may be of interest to reproduce laboratory notes relating to a 
marble which presented several problems in one — removal of an 
incrustation, removal of stains, consohdation, retouching, and repair. 
The marble in question is the head regarded as that of Mithras, from 
the Walbrook Mithraeum in London, excavated by Mr. W. F. 
Grimes in 1955 (Pi. 52). 

This fine head came from a moist site where it had been in contact 
with ferruginous clay, and it arrived at the laboratory damp, muddy, 
and encrusted with iron compounds. The stone, particularly in the 
region of the hair and peak of the Phrygian cap, was found to be 
weakened by a form of sugary decomposition, which might have 
resulted from the action of intense heat. There was no evidence to 
show that the temple had at any time been burned down, but a 
carbon residue was found in the incrustation, suggesting that the 
marble may have been exposed to altar fires. 

As the stone was considered to be too fragile for treatment in its 
damp condition, it was given time to dry out, and was then treated 
as follows: 

1. The bulk of the incrustation was removed with pointed match- 
sticks, exposing a surface that was porous and deeply stained in parts 
by iron compounds. Match-sticks were used because they are softer 
than marble and could not injure the surface. 

2. The iron stains were removed by sequestrating reagents, apphed 
to the surface with cotton-wool wrapped round the end of a stick. 
This was a long and painstaking operation, as it was only possible to 
work on one small area at a time. The most effective reagent was 
found to be Versenol,^ and, although this substance also sequestrates 

Distributors: F. W. Berk & Co. Ltd., Commonwealth House, 1-19 New Oxford 
St., London, W.C. i. 



calcium, it had no apparent effect on the marble in the dilute solutions 
employed. Care was taken not to carry this treatment too far. The 
discoloration resulting from the burning could not be obliterated, 
and it was therefore inadvisable to make the rest of the marble too 
white, as this would have caused it to have a patchy appearance. 

3. The sugary condition of the burnt areas was dealt vidth by apph- 
cations of lime water (calcium hydroxide), which, in time, was con- 
verted in the pores of the marble to calcium carbonate, a material of 
the same chemical composition as the marble itself. A small amount of 
crystalline efflorescence was brushed away, and, finally, two coats 
of a 10 per cent, solution of soluble casein were apphed to the areas 
of sugary marble to consolidate the surface, and the front of the 
Phrygian cap was hghtly touched out for aesthetic reasons with 
ground pastel colour. 

During the course of this work the neck of Mithras was exca- 
vated from the temple. It proved to be in a much better condition 
than the head and could be cleaned by ordinary washing. The neck 
and head were then joined together by a stout dowel of Delta metal’ 
inserted with plaster of Paris into opposing holes drilled in the frac- 
tured surfaces. 

The Mithras marble was only slightly burnt and was therefore 
amenable to treatment, but where marbles have been exposed to 
temperatures sufficiently high to convert the smrface into quickhme, 
there is nothing that can be done to restore them. Sometimes the 
heating may not have proceeded so far, and the surface may still be 
hard and intact though blackened vfrth tarry matter. Some improve- 
ment should be possible here, but it will depend on the nature of the 
specimen. The use of pyridine or morpholine, apphed in a coating 
of paper pulp, or mixed with an absorbent powder like kieselguhr 
and held in intimate contact with the stone until the solvent evapor- 
ates, may help to remove such staining. Ammoniacal benzene (see 
p. 309) may also help, but the final treatment may have to take 
the form of rubbing dovm the surface with water-of-Ayr stone or 
even with fine emery powder. Such treatment would obviously be 
' Delta Metal Co. Ltd., 295 Tunnel Avenue, London, S.E. 10. 



tmsuitable for fine sculpture, but it might be the means of saving a 
good piece if the stain were restricted to the plinth, or to a small part 
where the modelling was suggested, rather than executed with 
precision. But, in general, mechanical methods of cleaning marbles 
involving the use of abrasives can seldom be justified in museum 
work, and are only used under exceptional conditions. 

The use of copper chisels has been advocated as a means of cleaning 
marble gravestones and the hke. Since copper is of the same order of 
hardness as firesh marhle the procedure is not so drastic as it would 
appear to be. It is, nevertheless, quite imsuitable for museum work 
where surfaces are studied at close quarters and patina, an important 
feature of the specimen, must be preserved at all costs. 

The patina of marble. The patina of marble differs from that of 
metals in that it is essentially of the same chemical composition as the 
material upon which it has been laid down, whereas, in the case of 
metals, the patina is non-metaUic and generally consists of a variety 
of chemical compounds, oxides, chlorides, carbonates, &c. Marble 
patinas are formed in the same way as stalagmite by virtue of the fact 
that calcium carbonate is shghtly soluble in water that contains dis- 
solved carbon dioxide gas. When the solution (calcium bicarbonate) 
drips steadily from a height, the carbon dioxide gas is released, and 
the calcium carbonate is redeposited on the ground, where it may 
accumulate to form piUar-like structures or stalagmites. In the same 
way, when old marbles have been washed by rain for generations, 
there is a tendency towards solution and redeposition of calcium 
carbonate. Even though this be on a comparatively minute scale, 
a change of surface takes place and may be recognized as a variation 
in texture, translucency, or even colour, warm tones resulting from 
the presence of a httle ferric iron in the solution, and cooler greenish 
tones from ferrous iron or copper. Plants cause staining and organic 
growths such as hchens or algae may also leave a greenish tinge on 
the marble and the patinated surface is often of such porosity that it 
retains coloured impurities that add interest to the stone. Where 
patina exists, it is an essential feature of old marble and must be re- 
tained. It wiU survive cleaning by the general methods described, but 



would be ruined by the use of abrasives, sharp tools, or acids of any 

Examination of marble by U. V. fluorescence. It is instructive to exa- 
mine a piece of ancient marble in a darkened room by ultra-violet 
hght, and to compare the fluorescence of the patmated surface with 
that of a recent scratch or firacture in the same stone. Freshly exposed 
surfaces usually fluoresce brighdy, whereas the ancient surfaces are 
either deep brown or unchanged in appearance. It is often possible 
by examining a piece of marble sculpture under ultra-violet hght to 
determine whether it is recent or ancient and to reveal any modem 
cutting on an old marble, but as there can be no absolute guide as to 
the time required for a marble to acquire a patina recognizable as 
ancient, it is impossible to date marbles by the quahty of their 
fluorescence. Discretion is required in making such examinations, 
and, as the fluorescence of ancient marble is low in tone, it is essential 
to adapt the eye to a condition of total darkness for a few minutes 
before making the examination, so that sHght differences in fluores- 
cence can be discriminated. The problem is much simpler in the case 
of a composite piece made up from fragments of marble of different 
origins. An example is provided by a small Greek marble figure 
reclining against a tree trank. This appeared to be homogeneous but 
in a darkened room under the lamp was found to be heterogeneous; 
the torso, apparently original, was of one kind of marble, while the 
head and limbs were cut in two others, and the tree in yet a third. 
Even when the position of the joints had been discovered by this 
means, they were still very difficult to discern in dayhght. 

In detecting repairs to marble and estimating their extent, an ultra- 
violet examination is of the greatest value, and this type of investiga- 
tion is facilitated by the fact that plasters and glues fluoresce vividly. 
A valuable Itahan marble that had been lent to an exhibition in 
London was alleged to have suffered damage in transit assessed at 
a high figure. When examined by ultra-violet hght, however, 
the breakages were shown to be merely ancient repairs that 
had become unstuck on the journey. It was also revealed that on 
one part of the marble there was a shght abrasion (invisible in 


daylight) which fluoresced brightly and was obviously of modem 

It is perhaps not irrelevant to add that some of the modem surface- 
active agents used as detergents have a characteristic bright fluores- 
cence in ultra-violet hght, and hence it is not advisable to use them 
for cleaning marble that might have to be studied under the lamp. 
Detergents may have to be used for washing very dirty' neglected 
marbles, but traces of these materials left on or in the powdery 
surface of the stone would be likely to comphcate fluorescence 
analysis to such an extent as to vitiate this hne of inquiry. 

Effects of weathering. Numerous as are the problems of conservation 
of stone objects indoors, they are seldom so profound or so cmcial as 
those presented by stones exposed outside. These have to face ex- 
tremes of temperature and humidity, and the disintegrating action of 
firost, to say nothing of atmospheric impurities, and adventitious 
damage due to animals and plants. Sand or dust storms gradually 
erode the surface, and rain causes dirt to collect in hollows of the 
stone where stains are formed that can never be removed. 

A most striking example of weathering is provided by comparing 
photographs of Lord Elgin’s plaster casts of the West Frieze of the 
Parthenon made in 1802 (which show the condition of the marbles 
at that time) with photographs of the same marbles taken in situ in 
1938, i.e. 136 years later. The photographs show (Pis. 53 A and b) that, 
in what may be called the industrial age, the modelhng has every- 
where lost its sharpness, facial expressions have changed, and whole 
features have gone for ever, due not to the fault of anyone, but to 
the variety of causes that we know collectively as ‘weathering’. 

Marbles of all kinds are liable to be seriously damaged by the acids 
present in an industrial atmosphere. In the presence of moisture these 
destroy the surface, particularly at angles and comers, and open the 
joints to the disintegrating action of frost. It is not surprising, there- 
fore, that the industrial age has already taken a heavy toll of some 
of the finest marble monuments of antiquity. In the case of exposed 
buildings, the damage can only be arrested by the apphcation of 
stringent anti-pollution laws, but where statuary is concerned, a 



practical alternative may be offered by the possibility of covering in 
the marbles or bringing them indoors, so that they may be maintained 
under conditions where conservation is possible. 

Packing marble sculptures. The art of packing heavy stone objects 
can only be learnt by experience, but even the experience gained with 
sculptures of another kind may not suffice in the case of white marble 
sculpture, which is so susceptible to staining. White marbles should 
be packed in strong wooden boxes, made of weU-seasoned timber, 
assembled with brass screws. As the main requirement is to dispense 
with aU but the essential packing material, and at the same time 
ensure that the sculpture will not move in the case, wooden bars 
are screwed firmly in strategic places so that the marble is retained 
rigidly in position. The wooden bars must be padded with the best 
quahty white cotton-wool covered with tissue paper. The final 
retaining member should be a wooden wedge similarly padded, and 
driven home gently with a maUet to anchor the stone in position. 
The wedge should be retained, if possible, with a screw and be 
marked as the first member to be removed when th? marble is being 
unpacked. On no account should the use of newspaper, wood-wool, 
or straw be permitted. When a marble has been allowed to remain 
in its box over a long period it is desirable to inspect the padding and 
verify the tightness of the retaining wedge if removal of the box is 
contemplated, as the padding may have lost its resdience and the 
object may no longer be held securely in position. 


Adhesives that are quite satisfactory for the repair of small stone 
objects are inadequate where weights of i lb. and over have to be 
carried. In such cases the fractured surfaces are brought together 
with metal dowels or pins, and by this means the strain is no longer 
confined to the fractured surfaces but carried by the body of the 

The choice of metal for dowelling is important. Much damage has 
been done under the guise of conservation by dowelling stones with 
iron, which, on rusting, swells and opens old cracks and gives rise to 


new ones. Stainless steel may be used, but generally copper or one of 
its alloys is a better choice, one of the best and strongest for the pur- 
pose being Delta metal. As a rule the metal dowel should be of 
rectangular cross-section, and not so heavy that its insertion weakens 
the stone. The ends may be bifurcated or the edges nicked with a 
chisel to help the metal to key with the cementing material and form 
a rehable bond. In joints of this kind the dowels are made to do the 
work and the crack stopped afterwards with some inconspicuous 
filler. The procedure of fitting a metal dowel is briefly as follows: — 

Drill or cut a hole in one of the fractured faces of the stone at the 
appropriate angle, and to a depth of a httle more than half the length 
of the dowel, and of a diameter Just large enough to admit it. With 
the dowel temporarily in position, mark the opposite face, and 
sink a similar hole at the correct angle, so that the faces can be 
brought into correct register. (If a second dowel is required, pro- 
vision is made for it in the same manner.) The dowel has now to be 
fixed in one-half of the joint, and this is done by wetting the stone, 
filling the hole to a reasonable depth with a cement such as plaster of 
Paris, and pushing in the dowel, and, if necessary, wedging it firmly 
in position. It is allowed to set overnight. Next day the two parts 
are brought together and any necessary adjustment made to the hole 
that is to accommodate the projecting dowel. It now remains to 
wet the stone and apply the cementing material to the second hole, 
and it may be desirable to coat the surfaces as well with a thin wash 
of plaster to help to bed the joint in position when the pieces are 
brought together. When the joint is permanently set, then the crack 
has to be made good. This is done with some form of plastic stopping 
which may be a cement or a mastic, i.e. adhesive plus an inert sub- 
stance to act as a filler. 

The following cements may be used — plaster of Paris; Keen’s 
cement, which is a retarded plaster of a harder nature; Sirapite, 
a white dolomitic gypsum containing about i per cent, of free Hme; 
or Portland cement, the hard strong cement obtained by heating a 
mixture of clay and chalk. A suitable mastic may be prepared by 
grinding kieselguhr in a binder such as sihcate of soda or glue size. For 



use with antiquities, the best filler is undoubtedly plaster of Paris 
because of the ease with which it can be apphed and removed if this 
should ever be necessary at a future date. It is used for filling, and for 
repairing, and even for mounting heavy sculpture on stone plinths. 
When deahng with large plaster mixes the quick rate of setting be- 
comes a problem but this may be overcome by adding a htde glue 
to the plaster when it is made up. This has the effect of making the 
material workable for a longer time. 

For mounting and consohdation work (see p. 323) it sometimes hap- 
pens that a cement is required which is anhydrous and yet guaranteed 
not to shrink. This can be made from a 10 per cent, solution of celluloid 
cuttings dissolved in a solvent consisting of equal volumes of acetone 
and amyl acetate. If 50 grams of fine white sand are worked up with 
12 grams of this nitrocellulose solution, the mixture will be found to 
be of the consistency of soft putty which can be shaped with a palette 
knife. As it dries it adheres very firmly. It takes about four days for a 
block ^ in. thick to dry through, and then it has the appearance of 
light-coloured sandstone. A test block 4 in. by 2 in. by | in. made 
according to this formula showed neghgible contraction on drying. 


Clay is a hydrated aluminium sdicate of variable composition 
which has the fundamental characteristic of being plastic when moist 
so that it can be modelled or moulded; it hardens on drying, but can 
be softened again by the action of water. The nature of clay is changed, 
however, by the action of heat, so that it becomes rigid, stony, and 
almost imperishable vtithout change of form, and these quahties have 
been exploited by man in making bricks, terracotta objects, pottery, 
and porcelain. When baked above a temperature of 600° C. the 
change is irreversible, but if inadequately baked the material is of un- 
certain hardness and not necessarily permanent. There are thus three 
types of material presented for conservation — baked clay, unbaked 
clay, and clay that has been inadequately baked. In the latter cate- 
gories are the writing tablets of clay that are too frail to handle imtil 
they have been hardened by heating in a kiln. 



Clay tablets and their treatment 

Clay was widely used in Sumeria, Babylonia, and Assyria in the 
plastic condition for writing upon, cuneiform characters being im- 
pressed in the moist surface with a stilus. The clay soon regained its 
soHdity in the hot sun and there is evidence that, on occasion at least, 
it was further hardened by firing, but it is not known to what extent 
kiln firing was general. The hardness of the clay tablets that are 
excavated varies considerably and it seems likely that many were not 
kiln-hardened but merely sun-dried. Hundreds of thousands of clay 
documents have come down to us — the British Museum alone con- 
tains upwards of 100,000 — ^and they have survived astonishingly 
well, on account of the dry climate of Mesopotamia. 

But clay objects suffer in the same way as all porous material that 
has been buried in salty ground — they absorb salts which will tend 
to crystallize out after excavation and cause disintegration. The usual 
treatment for the removal of salts is washing, but with clay objects 
this is not possible as water would simply convert the clay into mud. 
As it is impossible to judge by inspection what heat treatment a cunei- 
form tablet may have been given in antiquity, the first requirement 
is a controlled rebaking of the old tablet in order to convert it into 
brick. The tablet may then be washed, or indeed soaked for pro- 
longed periods without being any the worse. The laboratory treat- 
ment of clay tablets thus consists in baking, washing, repairing, and 
cleaning the surface by mechanical methods should this be necessary to 
render the writing legible. Pis. 54A and B illustrate how, by this treat- 
ment, encrusted tablets can be strengthened and the writing made legible. 

Baking and washing clay tablets. In the British Museum baking is car- 
ried out in an oven or furnace in which the temperature is gradually 
raised to 750° C.^ and maintained for some hours. This results in the 
loss of what is known as ‘combined water’, and the conversion of the 
clay into brick. The tablets are allowed to cool in the closed furnace 

‘ P. Dclougaz (Oriental Institute of the Universit)' of Chicago, Studies in Ancient 
Onental Civilization, No. 7, 1933), in describing ‘The Treatment of Clay Tablets in the 
Field’ advocates a final heating at a temperature of 1,500-1,600° C., but such a high tem- 
perature would appear to be unnecessary and undesirable. 



overnight. When the baking has been properly conducted the tablets 
are converted to a clean hard material of a biscuit or pale ochre 
colour, depending on the amount of iron present in the clay. Any 
organic matter is, of course, burnt away, and in cases where there is 
saline impurity of a selenitic nature, this is dehydrated in the course 
of heating, leaving a white powder)' residue which may be removed 
by brushing. When, as is generally the case, soluble salts are present, 
it will be necessary to immerse the baked tablets in running water 
for about six hours to wash away aU but a negligible residue of soluble 
sahne matter. 

Unlike metal objects, where it is necessan' to wash away the last 
trace of salt because even a small residue might cause a renewed out- 
break of corrosion, stone, whether natural or artificial, is not affected 
by the presence of a shght amount of salt and may be considered safe 
as long as the bulk of the salt has been eliminated. 

Rendering cuneiform writing legible. To render cuneiform inscrip- 
tions legible, the old-fashioned methods of surface cleaning such as 
picking with a needle or treating the baked tablets with hydro- 
chloric acid are now out of date for routine purposes, although they 
may still be useful on occasion. These methods have been largely 
superseded by the sand-blasting technique. The sand or shot-blast 
method of cleaning^ is both speedy and effective; while removing 
debris and adherent particles from the inscription, it need not impair 
the surface of baked clay tablets in any way. A useful portable form 
of the apparatus has been devised in the British Museum Laboratory. 
This consists of a brass cylinder containing the sand or other grit. Air 
is blown through a tube below sand level, so that the particles form 
a suspension in the air stream and are ejected with the air from the 
nozzle which may be made from a curved glass or Polythene tube. 
Such an apparatus may be improvised from ordinary laboratory 
equipment, the air supply being obtained from a vacuum compressor 
at about 5 lb. per sq. in. pressure. The potency of the instrument will 
clearly depend on the nature and size of the particles, the shape of the 
jet, and the pressure and volume of the air. 

* The commercial form of shot-blast apparatus is referred to on p. 204. 

B 6157 




The apparatus illustrated in Fig. ii was designed for use in the 
field. It requires only a source of compressed air, which may be con- 
veniently obtained either from a cylinder or firom a motor-driven 
compressor. For cleaning tablets the jet should not be held too near 

FIG. II. Portable sand-spray apparatus with finger-tip 
control designed for field use 

the writing, nor should the pressure be allowed to become too strong; 
this is controlled by a movement of the finger over a hole in the feed 
pipe. The size of the jet must obviously be related to the coarseness 
of the sand, which is screened by passing it through a sieve of appro- 
priate mesh before it is fed into the cyhnder. This excludes the larger 
particles that would clog the jet. 

It should be added that while sand may safely be used out of doors 



(working down wind), it should never be used indoors for health 
reasons, as it is the cause of the disease known as sihcosis. Bauxite grit, 
however, may be used indoors with safety: it is an aluminium hydrate 
and cannot cause sihcosis though it must be admitted that the grit is 
just as unpleasant as aU other forms of dust, and the operator is well 
advised to do the work in a fume cupboard or where there is good 
ventilation. Garnet powders are also sold for grit spraying. 

Repairing cuneiform tablets. When a tablet is broken the fragments 
are baked separately, stuck together with a nitrocellulose adhesive 
such as Durofix, and then washed and dried. Occasionally one comes 
across cases where the clay is so crumbly that it would fall to powder 
if it were heated, and then the only course of action is to try and make 
good without baking, by consohdating and sticking the particles 
together. Tablets so treated may be further strengthened, if necessary, 
by filling cracks with a mixture of nitrocellulose and baked clay 
powder apphed as a thick paste. 

Consolidation of objects of unbaked clay 

In cases where clay cannot be hardened by firing, as in certain 
ethnographical specimens and in wall-paintings, consohdation is only 
possible by impregnation and for this purpose thin lacquers of poly- 
vinyl acetate, polymethacrylate, or nitrocellulose have given suc- 
cessful results. It is not advisable to use aqueous consohdating agents 
such as glue or casein. These might have a disintegrating effect on the 
clay in the first instance and, moreover, the water would be hable to 
affect colours adversely and to bring any soluble salts to the surface. 
Durofix may be used for repairs, mixed, if necessary, with fine clay. 
Such mixtures shrink on setting and the non-shrinking nitrocellu- 
lose/sand cement referred to on p. 319 is to be preferred. 

This cement was used for the surround and backing of Sir Aurel 
Stein’s large frescoes from Chinese Turkestan, now at the British 
Museum. The ground on which the frescoes were painted was mud 
and on arrival at the Museum the slabs were laid face downwards 
and the thick clay backing was cut away. About in. of the 
original groimd was left and this was reinforced either by a textile 



covering or by a sheet of expanded aluminium, the adhesive being 
nitrocellulose and sand. The frescoes that were backed with textile 
were then framed upon a sohd wooden backing, and then the sur- 
round and lactmae in the frescoes filled with the nitrocellulose cement, 
which had the colour and general appearance of artificial sandstone. 

The pieces that were mounted on expanded aluminium frames all 
belonged to one large fresco and were hung on the wall by attaching 
them to an armature that had been prepared for their reception. It 
was possible to set each slab in its permanent position in relation to its 
neighbours by moving pairs of screws at right angles. All that re- 
mained to be done was to make good the joints, using once again the 
nitrocellulose/sand cement which proved invaluable. 


The dark bituminous substance known as Eiimmeridge shale is easily 
carved and turned, and on this account it was used for making a 
variety of objects — spinning whorls, weights, bowls, and even furni- 
ture in the Roman period. It has deteriorated considerably with the 
passage of time, however, so that artifacts of shale are usually found 
in fragments or in such a condition that fragmentation soon follows 
excavation, and this is due to the shrinkage that accompanies drying. 

The form taken by the disintegrating shale depends upon how 
the shape and ornament are orientated relative to the bedding 
planes. When the two are roughly parallel, the shale sphts into a 
series of thin distorted leaves which may break up stiU further; but 
where the carving is at an angle to the bedding planes, a series of 
small crescent-shaped cracks appears, and these open up in the manner 
of a ripened fir-cone as the shale dries. Disintegration wdl follow 
unless the object is immersed in water; if this is done in good time, 
however, the cracks will close up and the surface may assume its 
original appearance. 

The only certain way of preserving all surface details is to keep an 
object of Kimmeridge shale continuously wet, and this can be done 
by sealing it in a jar of water containing about 15 per cent, of 



glycerine, to which about i per cent, of carbohc acid is added to 
act as a preservative and prevent the growth of moulds. 

In cases where it is considered desirable to dry the shale, and where 
a certain distortion of surface can be tolerated, the most effective 
procedure is to pass the object through a series of alcohols and then to 
impregnate in a vacuum with a consohdant such as an alcohoHc solu- 
tion of shellac. This, however, is not enough; the surface must be 
kept under continuous compression until the shellac hardens, and no 
further movement of the shale is possible. This is carried out as 
follows. The surface is first covered with absorbent paper (which can 
easily be removed at the end of treatment) and the object padded, if 
necessary, and bound tighdy with Polythene tape. If the shape allows, 
the shale is then placed in a carpenter’s vice provided with thick 
facings of soft wood or cork and allowed to dry for several days 
under pressure which is so exerted as to close the cracks; additional 
protection may be given to any protruding portions by the apphca- 
tion of shaped pieces of wood and the use of cramps. While such 
treatment prevents the loss of major fragments, it will be found, 
after removing the protective tape, that the surface is covered with 
a system of micro-cracks which have resulted from shrinkage. They 
are no great disfigurement as they are eventually filled with shellac 
during the final cleaning process. Ethyl lactate is recommended as 
the most effective solvent to use for the removal of excessive shellac 
at the conclusion of treatment. 

Objects of shale differ greatly in their appearance and behaviour 
after excavation, and a warning should be added that the restoration 
of important specimens should not be hghtly undertaken as the 
process of drying and consohdation may well extend over several 
weeks, and the work may require regular attention during this time 
if the results are to be satisfactory. 



It has already been noted that while clay is plastic in the moist 
condition, it can be hardened to a stone-hke mass on heating to red- 
ness. This hard condition persists on cooling, and no amount of 
soaking in water will soften the material again because a permanent 
and irreversible change has taken place. Baked clay is also highly 
resistant to chemical action and in this sense it is one of the most 
stable materials known, rivaUing gold itself in permanence. 

Raw clays vary in their chemical composition, and in the nature 
and quantity of the impurities they contain. It is only to be expected, 
therefore, that considerable diversity will be found in clay products. 
Examples of pottery of widely differing physical properties may 
even be made from one and the same clay, depending on the tem- 
perature and duration of firing and whether this has been carried 
out in the presence or absence of oxygen. Surface texture may also 
be varied in many ways. A common method is to dip the air-dried 
clay object in an aqueous suspension (slip) of an entirely different 
clay, prior to firing. Another method is to pohsh the surface before 
the clay is fired; this is possible because, in the normal course of air- 
drying, wet clay passes through a stage described by potters as being 
‘leather-hard’ and in this condition it can be turned on the lathe and 
burnished, so that when eventually fired the surface is smooth and 
has a dull polished appearance. Or again, the fired clay or biscuit 
ware may be glazed. 


Though it may be regarded chemically as an inert material, baked 
clay or pottery varies considerably in its physical properties. In- 



adequate baking yields pottery which is soft and porous and may be 
very frail in the wet condition. When excavated, pots are often ftag- 
mentary. The pieces require cleaning in order to remove earthy 
deposits and salts: they require strengthening, and in due course 
piecing together — ^with possibly some reconstruction of missing 
portions in plaster of Paris. 

Pottery is often found to be damp through and through when 
freshly excavated, ^ and in this condition it had best be allowed to 
dry before much is done in the way of cleaning. Pieces should be 
marked on the inside surface with waterproof Indian ink, and at this 
stage any interesting matter examined very carefully in case it should 
provide information that may be used in interpretation. Cleaning 
in the field should be delayed when the pot bears evidence of food 
remains or imprints of seeds as these will require to be submitted to 
expert examination. In other cases surface dirt may be removed by 
washing, adding, if necessary, a small amount of a peptizing agent, 
e.g. a surface-active agent such as Teepol^ that will tend to convert 
the mud from its normal condition of a colloidal gel into that of 
a sol; it then ceases to adhere to the pottery and can be washed away. 
Deposits of lime and chalk may remain, and before using solvents to 
remove them it will be necessary to determine the nature of any 
existing decoration and its abihty to survive cleaning. Soaking in 
dilute hydrochloric acid might be ideal for cleaning one type of 
sherd but would certainly not do for all. Thus, acid would be likely 
to ruin any pots that had been fired at a low temperature or even 
coarse terracotta excavated from chalky soil where the pores of the 
ware are filled with calcium carbonate. In this case the acid would 
react with the carbonate and the resulting effervescence might cause 
disintegration of the pottery. 

Unglazed earthenware is, of course, easily disrupted by salt, and 
Egyptian amphorae often display a broken external surface due to 
the cumulative effect of salt crystaDizations. Salt must be washed out 

' The excavation, recording, and repair of pottery is fully dealt with in R. J. C. Atkin- 
son’s Field Archaeology (Methuen & Co. Ltd.), 2nd ed., revised 1953. 

* Shell Chemicals Ltd., 105 Strand, London, W.C, 2. 


by soaking in changes of water and, after drying, the ware may be 
strengthened by impregnation with a solution of nitrocellulose. 

In cases where a pot has been inadequately fired, or where painted 
decoration is not adhering well, washing with water should be 
avoided unless the object has been previously consohdated with 
nitrocellulose. Ostraka are in this category. The writing is normally 
in a carhon ink, but this might easily be lost in the washing process 
unless it is first fixed by applying coats of a 2 per cent, solution of 
nitrocellulose in equal parts of acetone and amyl acetate. After this 
protective coating is quite dry it is then possible to wash the ostraka 
in the usual way to remove the salts without any fear of loss of 


As glazed pottery is covered by a vitreous film which is water- 
proof, it is much more permanent than unglazed ware. If, however, 
the glaze layer is incomplete or imperfect, soluble salts may get into 
the body of the ware; these will tend to cr^’staUize and cause the glaze 
to flake off. Although glazes exist in infinite variety, they only 
seem to give serious trouble when the attachment to the body is 

It is usually a lengthy and difficult business to extract the salt from 
the body if the only means of getting it out is through cracks in the 
glaze. This may be facihtated by immersing the object in dilute 
alcohol, which, though not a very good solvent for salts, will pene- 
trate better than water. The most effective treatment seems to be to 
apply the paper pulp technique, ^ see p. 299. This is not always possible, 
however, because the glaze may be loose, or fragments may be held 
to the ware only by the salt crystals. In such cases it is necessary to 
fix the glaze hy applying a dilute solution of nitrocellulose dissolved 
in equal volumes of amyl acetate and acetone, and, later on, to wash 
out the salt in changes of distilled water. Some further consohdation 
will no doubt be necessary after the salts have been removed. 

' Penetration will be facilitated by the addition of one volume of Lissapol N to 200 
volumes of distilled water. 



In common with certain kinds of enamel and glass, glaze is subject 
to devitrification — a loss of translucency caused by crystallization or 
deposition of one or more of its ingredients. When sUica is deposited 
in quantity, the glaze will become opaque or it may take the form of 
an opalescent film veiling the colour and ornament. In the case of 
certain pieces of Persian pottery it was possible to restore the trans- 
parency and to reveal underglaze decoration by local apphcations 
of a I per cent, solution of hydrofluoric acid for a few minutes at a 
time, washing with water, and rubbing the treated areas afterwards 
with wet glass-paper of medium grade. When the opalescent deposit 
is confined to the immediate surface, as is usually the case, this treat- 
ment restores transparency. The cloudiness may go deeper, however, 
or it may be deposited irregularly, and in this event it may be unsafe 
to attempt to do more than treat the surface. If by such treatment it 
is not possible to clear the surface in the first instance, it may be that 
lacquering will restore transparency^ and for this purpose any clear 
lacquer of high refractive index will suffice. 

A common feature of some Chinese pottery is a crackled glaze, 
and this might seem at first sight to be a defect, but this is very far 
from being the case. Crackle may arise from many accidental causes, 
but in the long run it is due to differences in the relative contraction 
and rate of coohng of the glaze and the body. The Chinese potter 
was able to control conditions so as to produce a definite artistic 
crackle and thus make capital out of an apparent deficiency. 


Potter)^ earthenware, tiles, and suchhke are all porous, and if 
they require to be strengthened, this is done by impregnation, using 
dilute synthetic lacquers containing polyvinyl acetate or polymeth- 
acrylate. In some cases, vacuum impregnation may be desirable (see 
p. 304). Repairs may be carried out with Durofix and, if necessary, 
a joint may be strengthened by dowelhng (see p. 317), or by attach- 
ing a bandage along the back where it -will be inconspicuous. 

It is not possible to repair dusty joints. The first step, therefore, 
consists in getting rid of dust; then, if the ware is not very absorbent. 


the inside of the fractured edge is roughened in order to give a key 
for the adhesive. The next step consists in painting adhesive on both 
edges and this is allowed to become tacky. A thin film of fresh adhesive 
is then apphed to the edges, and the joints are pressed firmly together. 
Pressure should be maintained for at least halt an hour by an elastic 
band or by some other means to ensure that satisfactory joints are 
formed. In this connexion a sand-box is a great convenience as the 
repaired pieces may be stuck in the sand temporarily at any appropri- 
ate angle, so that joints will not be under strain whilst the adhesive is 
hardening. Small muslin bags of sand are also useful to support larger 
and heavier pieces of pottery in the course of reconstruction work. 
When dry, any excess of adhesive that has been squeezed out fiom 
the joint may be removed Avith a sharp chisel, or even by rubbing 
with a cloth moistened with a httle solvent, though, if the latter ex- 
pedient is adopted, care must be taken that there is not enough sol- 
vent present to run into any cracks and soften the joints. 

The following adhesives may be used as alternatives to Durofix. 
For large work, a retarded plaster, e.g. ‘white glue’ (hot glue to 
which has been added plaster of Paris); for dark pottery in which re- 
paired joints may have to be softened by heat in the course of making 
a compHcated reconstruction, thick shellac in spirit or even sohd 
flake shellac apphed with the aid of a tiny flame. 

For making permanent joints that are very strong and waterproof 
an epoxy resin adhesive, e.g. Araldite ioi,i is recommended. In using 
this adhesive, the pieces must be heated, preferably to at least 
i8o° C., and the Araldite then apphed to the hot joints like sealing 
wax or as a powder according to which is the more convenient; 
the pottery or porcelain is wired, or otherwise held firmly together, 
whilst the Araldite is polymerized in the oven. The cured adhesive 
has a brown colour but is very strong and rehable. 

Another adhesive, marketed by the same firm, and which can be 
specially recommended for unglazed ware, is the resorcinol-formalde- 
hyde resin known as Aerodux 185. This is a cold-setting, gap-filhng 

' This is marketed by Aero Research Ltd., Duxford, Cambridge, in the form of sticks 
or as a powder ready for apphcation. 



adhesive resistant to boiling water; it is mixed with a hardener before 
being appHed to the fractured surfaces. The fact that it sets in the 
cold is a great convenience for many types of objects; and, since it is 
a gap -fillin g cement, it has the advantage that the joints which are 
being fitted together need only be Hghtly clamped. See also p. 13 1. 

In the reconstruction of pots from broken fragments, it often hap- 
pens that pieces are missing, and it may be desirable to fill the lacunae 
with plaster that has been coloured to match the pot. The procedure 
is as follows. If the missing portion is triangular, clean the three sides 
and, if the pottery is finely grained, score the edges longitudinally in 
order to give the plaster a key. Soften some Plasticine by warming 
and rolhng it on a glass plate, and apply a slab, a quarter inch thick, 
to the inside of the pot covering the missing portion, so that, when 
the space is filled with soft plaster, the Plasticine will form a backing 
and reproduce the correct curvature. With narrow-necked vessels 
this is not possible, and the plaster must be built up from outside, 
little by little, to fill the space without any backing; the repair will 
bulge shghtly on the inside where it is not seen. 

Before applying plaster it is necessary to stop the suction, i.e. to 
treat the pottery with a Uquid in order to prevent it absorbing mois- 
ture from the plaster and thus reducing its cohesion. This could 
be done with water, but pottery varies so much in its condition — 
strength, porosity, &c., particularly when it has been broken in 
pieces — that water is better avoided in this kind of repair work. A 
dilute solution of shellac is much more effective. The joint is painted, 
therefore, with thin bleached shellac or French polish diluted with 
methylated spirits. 

Coarse baked plaster of Paris is used as a rule unless for exception- 
ally fine work when boiled plaster is the grade preferred. ^ 

If a coloured plaster is required for repair work, dry powder 
colours are added to the plaster powder before it is made up with 
water. These colours should be mixed well with the plaster on a glass 
plate, using a steel spatula. When water is added the colour will 
appear much darker, but on evaporation of the water the plaster 
* CafFarate and Co. Ltd., Newark on Trent. 



will gradually recover the original hue that was estabhshed before 

To mix the plaster, put a quantity of cold water in a bowl and 
shake the powdered plaster on to the surface until it stands as a 
conical heap above the level of the water. When air bubbles cease 
to rise, stir with a spoon imtil a thick cream is obtained that shows 
signs of becoming thicker within half a minute, when it is ready to 
be appHed to the break. If the plaster fails to thicken quickly in the 
bowl, do not stir violently, but add some more plaster powder until 
the correct consistency is obtained before using it to repair the pot. 

When the pot has been prepared and the plaster mixed, it may be 
filled into the gap with the aid of a spoon. Avoid imprisoning air- 
bubbles and continue filling the gap until the fresh plaster is every- 
where shghtly above the surface. After a few minutes the plaster will 
have set to a buttery consistency and it is then an easy matter to pare 
it down with a plaster scraper (a metal spatula having a comb-hke 
edge) until it is level with the surface of the pot. After half an hour or 
so the surface may be finished with the appropriate grade of glass- 
paper — o, 1 ^, or 2, as the case may be. The Plasticine is then removed 
from the back of the repair and the exposed surface of the plaster 
made good as required. 

If, after drying, it is found that the colour of the repaired area does 
not match the pot, the whole area of new plaster will have to be 
stopped again with a thin coating of white French poUsh before re- 
touching with fresh colour. This is essential, because, if colour is 
appHed without previous stopping, the binding medium will be 
sucked from the fresh colour into the body of the repair, leaving a 
‘watermark’ stain which is difficult to remove. To prepare a matt 
paint for retouching pottery, the pigments — ochres, umbers, black, 
white, or green — should be mixed and ground either in skimmed 
milk or in dilute size solution. 


Coated \\ 1th paper pulp to remove soluble salts 



Glass results from the fusion of acidic and basic oxides, the chief 
acidic oxides being silica and boron oxide, and the chief basic oxides 
soda, potash, lime, alumina, htharge, and magnesia. When selected 
mixtmes of these oxides are fused at a high temperature, a clear 
hquid is obtained, which, on coohng, becomes a transparent amor- 
phous sohd. Thus, soda-hme glass is made by fusing soda, hme, and 
clean sand (= silica) in a refractory crucible. This is the commonest 
type of glass. 

The physical properties of glass may be varied according to the 
nature and proportions of its constituent sihcates. For example, 
potash-Hme glass (crown glass) has a high fusion point. Potash-lead 
glass (flint glass) ^ is softer and more easily cut and it has, moreover, 
a high refractive index, and is sometimes referred to as ‘crystal’. 

In the molten condition, glass is an excellent solvent for metaUic 
oxides, and some, such as cobalt oxide, copper oxide, and iron oxide, 
are used to give the glass characteristic colours while retaining trans- 
parency; others, such as tin and antimony oxides, cause opacity and 
impart to the glass a white appearance. Opacity results also from the 
use of larger quantities of colouring oxides. 

Until the seventeenth century, almost all glass was composed of 
sihca, hme, and soda or potash. The glass-makers themselves thought 
of their material as being made from only two constituents. Neri^ 
defines glass as ‘a concrete of salt and sand or stones’. The salt, at 

' The presence of lead in glass may be detected by applying a drop of hydrofluoric 
acid in an inconspicuous place and after a moment absorbing the drop in filter paper that 
has been previously moistened with ammonium sulphide solution. A black stain on the 
paper indicates the presence of lead. 

^ Neri, A., The Art of Glass, pubhshed in Florence, 1612. Merrett’s translation, 1662. 



least in Europe, was derived from plant ashes, but plant ashes con- 
tain much besides the alkah salts; they contain lime, alumina, &c., 
and because of this a stable glass was produced. This early glass was 
generally coloured dark green by the presence of iron oxide, and 
unfortunately, in their desire to produce a pure white glass, some 
makers used salts leached from the ashes and produced a glass danger- 
ously low in hme. This was noticed by Neri, who wrote : ‘furthermore, 
in the finest Glasses, wherein the salt is most purified, and in a greater 
proportion of salt to the sand, you shall find that such Glasses stand- 
ing long in subterraneous and moist places will fall to pieces, the 
union of the salt and sand decaying’. 

If the early glass-makers had been using pure ingredients, their 
product, potassium or sodium sdicate (known as water-glass), would 
have been soluble in water. Other ingredients are essential to make 
a stable glass. Of the sdicates that are combined to produce soda-lime 
glass, sodium sdicate is soluble in water and readily fusible, calcium 
sdicate is insoluble and fuses with difficulty; but a correctly balanced 
mixture of the two combines to form a glass that is insoluble and fuses 
at a moderate temperature. 

In an interesting paper on the composition and working properties 
of ancient glasses, Matson^ states that ‘The eutectic mixture in the 
soda-lime-sdica system is, approximately, sdica 73 per cent., hme 
5 per cent., soda 22 per cent.’, and he advocates using this as a yard- 
stick in studying analyses of ancient glasses, some of which he quotes. 
By studying the complete analysis, he shows that one can hazard a 
guess at the working properties of the glass, e.g. excess of sdica makes 
the glass very difficult to work. But what is more important from 
the point of view of conservation, the analysis reflects also upon the 
stabdity of the glass, excess of lime tending to promote devitrifica- 
tion and excess of soda tending to make the glass susceptible to 
attack by water. It is not only the presence of certain ingredients 
but the proportions in w^hich they are present that determine the 
stabdity of the material. 

In considering the stabdity of glass there is yet another factor to 
* Matson, F. B.., Journal of Chemical Education, 1951, 28, p. 82. 



be taken into account. From the physico-chemical point of view, 
glass is regarded as a state of matter rather than a specific material. 
It is a supercooled hquid and therefore a metastable substance, and it 
is characteristic of material in this condition that cryst allis ation may 
take place at any time. Glasses having a balanced formula are remark- 
ably stable substances, but where injudicious changes are made there 
is a tendency for one or other of the ingredients of glass to crystal- 
Hze out, and when this happens the condition is known as devitri- 
fication. As we have seen in the case of glazes, devitrification causes 
opacity and opalescence; a partial devitrification may sometimes 
enhance the appearance of the glass, particularly striking effects of 
iridescent colour resulting from the diffraction of Hght by the 
thin films or flakes of which the surface of devitrified glass is com- 

Deterioration of glass 

The decomposition of glass is generally accompanied by the Hbera- 
tion of firee alkali, and the form taken by the disintegration will be 
determined by the nature and amount of the alkali hberated. Free 
alkali is more or less hygroscopic, and liberation of Kme and soda 
will cause deposition of moisture. Carbon dioxide will be absorbed 
from the atmosphere by this moist alkah, with the result that an in- 
crustation of alkahne carbonates is gradually laid down, interspersed 
with siHca, and with tiny flakes of semi-decomposed glass, the result 
being the creation of a surface that is opalescent. 

An opalescent surface requires no special treatment as it is stable 
under dry conditions. Unfortunately such a surface is frail and the 
glass itself brittle, and as any attempt to fix the flaking particles with 
lacquer might destroy the colour effect, lacquering should in general 
be avoided.* To remove dirt, such glass may be washed with a deter- 
gent and dried, preferably through baths of alcohol and ether. 

When much alkah is present, conditions are very different. 

' Nevertheless, remarkable effects have been achieved by using a special polymethacry- 
latc lacquer under controlled conditions as described by Hedvall, J. A., et al., Chalmers 
Tikniska Hogskolas Handlingar, 1951, Nr. 118. 


Potassium salts, for example, are much more hygroscopic than those 
of hme and when present in excess no incrustation forms, but instead, 
beads of potassium carbonate solution of a strongly alkaline nature 
rim down the glass and hasten its decomposition. Glass undergoing 
such decomposition is known as ‘sweating’ or ‘weeping’ glass and, 
if left untreated, will soon deteriorate in appearance and in time be 
destroyed completely. So long as the glass is damp it may remain 
transparent; unfortunately when the hygroscopic salts are removed 
and the glass is dry, the transparency is lost and can never be re- 
stored completely even by lacquering. 

‘Weeping’ glass is treated as follows : after washing in running water 
for a few minutes, place in a bath of sulphuric acid (2 per cent.) for 
several days to remove free alkali. Wash it once more and then 
dry, preferably through alcohol and ether. This treatment retards 
disintegration but can offer no permanent cure. Glass hable to 
‘sweat’ must be kept under reasonably dry conditions. Individual 
pieces may be kept in airtight boxes supphed with enough sihca gel 
to keep the contents dry. Such boxes can be constructed from Perspex 
and sealed at the joints with Polythene tape. As an alternative, an 
exhibition case may be adapted for the purpose as shown in Fig. 9. 

Much ancient glass must have been lost to us as a result of decom- 
position arising from an excess of Hme in the formula or from im- 
perfect mixing of the ingredients, due to inadequate heating during 
manufacture. In the case of opaque glass having what appears to be 
a stable surface, any heterogeneity in the interior will be concealed 
from view so long as the outer shell is intact. Such an exceptional 
condition was illustrated in the case of a large Egyptian scarab of the 
eighteenth dynasty that had been made by casting opaque blue glass 
in a mould. The scarab had remained intact for 3,000 years; it had 
been on exhibition in the British Museum for a generation, and there 
had been no reason to regard it as being in any respect abnormal. It 
was reported one day to have developed cracks, and was found to be 
so hygroscopic that it would not remain dry for two minutes on end. 
Soon afterwards it fell into about a dozen pieces. From an examina- 
tion of the interior of the scarab it was obvious that the condition of 

JZ. rili; MITHRAS 11I;A]) IROM W A I, IS R () () K , LONDON 

Lcfl: Before treatiiieiit. B. R{fihl: After eleaiiing anti reuniting with tlie neck, later tliscovered on 

the same site 



the inside was different from that of the exterior — a white incrusta- 
tion formed on parts of the fractured surfaces but no such crystalhza- 
tion took place on the outside. It only required a micro-crack to form 
in the outer shell for moismre to have access to the hygroscopic 
material inside, and when this happened disintegration was inevitable. 


An enamel is a lead glass to which some metaUic oxide has been 
added to colour it or make it opaque. It is apphed thinly to a metal 
base, usually of gold or silver. When glass is melted on metal, as in 
the manufacture of enamels, the difference in the degree of contrac- 
tion on cooling between the glass and the metal is considerable, and 
sometimes a state of strain is set up between the two. In order to 
equalize the strains, the enamel is sometimes apphed all over the back 
of the metal as well as the front. 

That the strains are not always equalized is evident from a singular 
occurrence that took place in the Wallace Collection in 1929. ^ 

An eighteenth-centrury enamelled snuff-box on exhibition in the 
museum suddenly disintegrated, scattering glass powder throughout 
the case. When the powder was examined in the British Museum 
Laboratory the particles were shown to be of the same shape as those 
formed by the bursting of Prince Rupert drops, that is, rounded and 
not splintery. These drops, which have been known for centuries, are 
pear-shaped blobs of glass made by pouring molten glass into water. 
Though apparently quite stable, and so strong that they can even be 
hit at the thick end with a hammer vrithout breaking, they are, in 
fact, imder great internal strain, and have the astonishing propert)' 
of flying to powder when the smallest fragment is broken from the 
fine glass filament forming the stem of the pear. When made in the 
form of a phial it suffices to drop a tiny grain of sand inside for 
the glass to fall in two, and it is likely that some such simple trigger 
action caused the enamel in the Wallace Collection to ‘explode’. 

Sometimes strain causes an enamel to become fractured and de- 
tached from the base, and when this happens the only thing to do is 
’ Camp, S. J., Art Work, 1930, 21, p 47. 



3j8 siliceous and RELATED MATERIALS 

to attempt consolidation by flooding the area with dilute lacquer so 
that it seeps through the cracks. When it has completely set, the 
excess of lacquer may be removed. In the coarser type of enamel 
work it is sometimes possible to execute repairs in coloured cellulose 
paints or varnishes that have been suitably dyed. 

A curious effect occasionally to be observed is that of fungus 
apparently growing on one specific colour — say on a translucent 
crimson. On closer inspection it is found that the nutrient material 
is gelatine, which has been used to reinforce the weak areas of the 
enamelling and possibly to act as a base for retouching with pigment. 
The treatment in this case is to remove the gelatine with warm water, 
sterilize the area with Santobrite (2 per cent.), and after drying con- 
sohdate with lacquer tinted as required. 

The opaque red glass of sealing-wax colour, often referred to as 
cuprous enamel, which was of universal use m the Bronze Age for 
the decoration of metal, is often found to be oxidized to a green 
colour. The earhest examples that have come to the British Museum 
Laboratory are from a glass furnace excavated at Nimrud (650 b.c.) 
and are in the form of lenticular cakes firom the crucible, some 8 
inches in diameter, and up to 2 inches in their greatest thickness. 
These are invariably green on the surface, and are like corroded 
bronze in appearance. The green layer is porous and extends to a 
depth of almost i/i6th inch, the interior being apparently in a perfect 
state of preservation, rich purplish-red, opaque, and having a charac- 
teristic glassy firacture. Unfortunately, in dealing with inlays of this 
red glass, the only means of revealing the red colour is to grind or 
scrape away the superficial green layer. This treatment is by no means 
always practicable, but has been successfully carried out in the case 
of a number of Saxon hanging bowls decorated with escutcheons 
in which this red glass was used for enamelling. The removal of the 
green film enhanced the appearance of the enamel considerably. 
When the glass is unattached to metal, or where an inlay can be 
removed for cleaning, it might be worth while trying to dissolve the 
green surface layer with a dilute solution of hydrofluoric acid. Pre- 
cautions are necessary, however. Immersion should be for short 



periods only; the work is inspected at frequent intervals by wash- 
ing under a running tap and the acid treatment stopped before aU of 
the green has disappeared. When washing with hydrofluoric acid a 
plastic photographic dish should be used, and rubber gloves should 
be worn to protect the skin, particularly under the finger-nails, where 
it is very sensitive to this acid. 

Repairing glass 

Broken glass is one of the most difficult materials to repair. If 
thermo-setting adhesives can be used, Araldite loi gives excellent 
results and it has actually been used to mend the broken stem of a wine 
glass. But it is seldom possible to take the risk of heating old glass 
which, if devitrified, would almost certainly disintegrate. Durofix 
is probably the most useful general-purpose adhesive, and another 
good one is made by dissolving chips of Perspex in the following 
mixed solvent, to give a solution of 10-15 per cent, strength; 

Ethylene dichloride . . . .195 ml. 

Glacial acetic acid .... 5 ml. 

Acid solutions of glue may also be recommended, and a rehablc 
adhesive can be made by dissolving Scotch glue (20 grams) in 30 ml. 
of hot glacial acetic acid to which i gram of ammoniiun bichromate 
is added. The glue is warmed in the usual fashion before apphcation. 

Adhesives are almost useless for the repair of glass that is very thin 
or much fractured. Where fragments of a glass vessel are in such a con- 
dition that its shape can be recovered, a transparent Polythene tape 
may be used, applying it to the inside, and beginning the reconstruc- 
tion preferably at the mouth of the vessel. This technique, however, 
requires dexterity and leaves much to be desired. It must be admitted 
that in this category of repair work there are cases that are apparently 
beyond recovery. 




Liquids dissolved in liquids are measured in fluid ounces per pint or 
in millilitres per litre: i.e. volume/volume, (v/v.). 

SoHds dissolved in liquids are measured in ounces avoirdupois per 
pint or in grams per litre: i.e. w^eight/volume, (w/v). 

10 per cent, v/v solutions contain 2 fluid oz./pint or 100 mls./Htre. 

10 „ „ w/v ,, „ 2 avoir, oz./pint or 100 gms./litre. 

Acids and alkalis. Concentrated acids and alkalis in liquid form are sold 
by specific gravity as they are not necessarily of 100 per cent, strength, 

Sp. Gr. 

Sulphuric acid . . . .1-84 

Nitric acid . . . . i -42 

Hydrochloric acid . . . i-i8 

Ammonia .... 0*88 

The convention adopted in this book, in preparing dilute solutions, 
is to assume that the commercial chemical is 100 per cent, and dilute 
accordingly. Thus, 

I per cent, nitric acid is made by dissolving i fluid oz. of nitric acid 
(i '42) in 5 pints distilled water. 

10 per cent, sulphuric acid is made by dissolving 2 fluid oz. of sul- 
phuric acid (f84) in i pint of distilled water. 

50 per cent, ammonia is made by mixing equal volumes of ammonia 
(o-88) and distilled water. 

Hydrogen peroxide 

This is available, commercially, as follows: 

‘100 volumes’ = approx. 30 per cent, w/v hydrogen peroxide, 
sometimes called Perhydrol. 

‘20 volumes’ = approx. 6 per cent, w/v hydrogen peroxide. 

‘10 volumes’ = approx. 3 per cent, w/v hydrogen peroxide. 
These solutions are supphed in coloured bottles and should be stored 
in a cool place protected from Ught. 



3. Enzymes 

Enzymes are naturally-occurring organic substances that have a specific 
action in catalysing many organic reactions, including the breakdown 
of sugars, fats, and proteins. Mixtures of enzymes are obtainable, 
called ‘digester powder’, and this may be used for rendering certain 
organic stains soluble, particularly those occurring on textiles (see 
p. 105), so that they may be removed subsequently by washing. 
Enzymes act best in weak solutions at temperatures between 37 and 
50° C. — they are soon destroyed at temperatures in excess of 60° C. 
The hydrolytic action of enzymes takes time to complete, perhaps 
an hour or longer, and the process is known as digestion. 


Although many chemicals may be regarded as potentially dangerous, 
special care is needed when using the following: 

(1) Concentrated sulphuric acid 

This can cause a serious bum if brought in contact with the skitu In 
preparing dilute solutions of sulphuric acid the concentrated acid should 
always be added slowly to the water with continuous stirring to dissi- 
pate the large amoimt of heat evolved. Never add water to the con- 
centrated acid. 

(2) Caustic soda 

The sohd should not be allowed to come in contact with the skin. In 
preparing solutions of caustic soda there is a considerable evolution of 
heat which may cause a thick glass vessel to crack. Such solutions are 
therefore best made by slowly adding the sohd (with care to prevent 
splashing) to water contained in a porcelain dish or an iron or stainless 
steel vessel standing in a sink. 

(3) Concentrated nitric acid 

Strongly corrosive and should not be allowed to come in contact with 
the skin, which it stains a deep yellow. 

(4) Concentrated hydrochloric acid 

Care should be taken when opening a bottle of this acid, as fumes are 
immediately evolved which will cause pain if they come in contact 
with the eyes. 

(5) Hydrofluoric acid 

Strongly corrosive. Usually handled in Polythene bottles. The fumes 
are very unpleasant if inhaled and vrill damage any glass with which 
they come in contact. 

(6) Formic acid. Carbolic acid [Phenol) 

Powerful skin irritants. 



(7) Concentrated ammonia 

Care should be taken when opening a bottle of concentrated ammonia; 
the stopper may fly off due to pressure. The bottle should be kept in a 
cool place and must never be allowed to remain in the sun. 

( 8 ) Hydrogen peroxide ‘100 vols.\ Perhydrol 

This must be kept in a cool place as great pressure may be set up in the 
bottle if left in the sun or near a source of heat. 

(9) Diethyl ether [B.P. 34-6° C.) 

Highly volatile and inflammable. Forms explosive mixtures with air. 

(10) Carbon disulphide (B.P. 46° C.) 

Highly volatile and inflammable. Poisonous. Forms explosive mixtures 
with air. 

N.B. A foam fire extinguisher should be at hand in any room where 
ether or carbon disulphide are used. 


(1) Acid burns. Flood the affected area with water and then wash with a 
dilute solution of sodium bicarbonate. 

(2) Caustic soda burns. Flood the affected area with water and then wash 
with very dilute acid, e.g. vinegar. 


Adhesives for paper should be free from staining material and harmful 
preservatives, and while the best photographic mountants may satisfy 
these conditions, cheap office pastes and adhesives supphed for hanging 
waU-paper should be avoided as well as all preparations of doubtful or 
unknown composition. 

For general work there is nothing to surpass freshly prepared flour paste 
(bookbinder’s paste), as it is easy to make and to apply and has excellent 
adhesive properties. It is made from ordinary (not self-raising) flour and 
water in the following proportions — 

White flour (wheat) . 500 grams 

Water .... 2-5 htres. 

Mix the flour with a Httle of the water in an aluminium or enamel pan, 
the lumps being broken up with the hand to form a smooth cream. Boil 
the remainder of the water separately and add it to the cream, stirring 
continuously. The paste is now heated, not directly, but by standing 
the pan in a container of water kept boiling (double pan. See Pi. 55). It 
should be stirred meantime and will soon thicken. After about ten minutes 
it may be decanted into a suitable vessel, and to prevent a crust forming 
as it cools place a sheet of paper on the surface of the paste and pour a Httle 
water above the paper. The mixture as prepared is too thick for general 
use: small quantities should be removed to a pasting dish and thinned 
with water as required. The paste keeps well for a few days in cool sur- 
roundings but should be discarded on the first sign of souring. 

If it is desired to keep the paste for as long as a week, about 10 ml. of 
formalin may be stirred into the mixture while fresh, but for ordinary 
use no other form of preservative can be recommended. Alum is often 
present as an ingredient in formulas for paste; it hardens the adhesive so 
that it tends to become insoluble in water, but alum has been shown to 
be harmful as it makes the paper acid (see p. 52). For use in the tropics, 
however, the paste must contain a substance that will inhibit the growth 
of mould and render it unattractive to insects, and for this purpose 
poisonous material is introduced. Suitable formulas have been suggested 



containing small quantities of phenyl mercuric acetate, or phenyl mer- 
curic borate, say, about half a gram for every 2J Htres of paste. 

In applying paste it should be spread evenly and thinly so that it will 
dry quickly, and there should never be so much water present as to cause 
the paper to stretch appreciably. The paste may be appHed with a brush 
or with an ivory or bone paper-knife. After pasting, the paper should 
be dried in the press or under a sheet of heavy plate glass. 


y = A-B 


(1) How much water has to be added to a 12 per cent, solution to make a 
5 per cent, solution? 

Let A = 12 per cent. 

B = 5 per cent. 

C = 0 per cent, (water) 

Then x = B—C = 5—0 — 5 parts of the 12 per cent, solution [A) 
y = A—B = 12— 5=7 parts of water (C) 

Answer: Mix 5 parts of the 12 per cent, solution with 7 parts of water. 

(2) How much of a 45 per cent, solution has to be added to a 17 per cent, 
solution to obtain a 30 per cent, solution? 

Let A = 45 per cent. 

B = 30 per cent. 

C = 17 per cent. 

Then x = B— C = 30— 17 = 13 parts of the 45 per cent, solution {A) 
y — A—B = 45—30 = 15 parts of the 17 per cent, solution (C) 
Answer: Mix 13 parts of the 45 per cent, solution with 15 parts of the 
17 per cent, solution. 



An acid may be defined as a substance which forms hydrogen ions when 
dissolved in water, and an alkah as a substance which forms hydroxyl ions 
when dissolved in water. Acids and alkalis are capable of neutralizing one 
another to form salts. 

When dealing with small concentrations of acids and alkaUs it is 
cumbersome to express the actual concentrations of hydrogen and 
hydroxyl ions in terms of weight per volume. A convenient method was 
proposed in 1909 by Sorensen who introduced the hydrogen ion exponent, 
commonly known as pH. The relationship between pH and hydrogen 
ion concentration is shown in the following nomograph: 

HYDROGEN JON id* lO’* lO'^ > 6 * 16* »6^ to^ o® o® o'® lo" k 5 ‘^ 











pMO I 2 3 4 5 6 7 8 9 OM l 2 13 14 


The advantage of the system is that very small concentrations of acid or 
alkali can be expressed as whole numbers. A solution in which the hydro- 
gen and hydroxyl ion concentrations are equal is an exactly neutral solu- 
tion and the pH = 7. In an acid solution the hydrogen ion concentration 
exceeds that of the hydroxyl ion and the pH is less than 7, whilst in an 
alkaline solution the reverse is the case and the pH is greater than 7. 

Certain substances known as indicators show characteristic colour 
changes depending upon the pH of the solution to which they are added. 
Litmus, for example, is red in acid and blue in presence of alkalis. Its 
pH range is 5-o-8'0. A more precise estimation of the acidity or alkalinity 
of a solution may be made by adding to it a few drops of Universal 
Indicator^ which shows graded colour changes from red to violet through- 
out the pH range as follows; 

' Indicators of this type are B.D.H. Universal Indicator obtainable from British Drug 
Houses, Poole, Dorset, and B.T.L. Universal Indicator obtainable from Messrs. Hopkin 
& Wilhams Ltd,, Freshwater Road, Chadwell Heath, Essex. 



Colour of indicator 

Nature of solution 

pH value 


Very acid 


Orange . 

Moderately acid 


Yellow . 

Slightly acid 






Shghdy alkaline 


Bluish-green . 

Moderately alkaline 



Very alkaline 



Intensely alkaline 



100 ° C. 

































212° F. 


































66 ° C. 


































150-8° F. 



































32° C. 

































89-6° F. 





























Temperature and relative humidity readings are taken with a thermo- 
meter and hygrometer respectively. Several types of hygrometer are 
available^ the most convenient being the sling or whirling hygrometer 
or psychrometer which consists of wet- and dry-bulb thermometers fixed 
side by side on a frame for s-winging, the wet bulb being kept damp by 
a wick dipping into a reservoir containing distilled water. To use the 
instrument it is held well away from the body and is whirled vigorously 
for at least a minute, readings on the wet-bulb and dry-bulb thermo- 
meters being taken immediately. Further observations should be taken 
till constant readings are obtained. The wet bulb will generally record 
a temperature lower than the dry one. This difference represents the de- 
pression in degrees resulting from evaporation of water, and from this 
depression, taken in conjunction with the temperature of the dry bulb, it 
is possible to determine the relative humidity by consulting hygrometric 
or psychrometric tables.^ 

The sling hygrometer has the advantage that it may be used in 
different parts of the room, and it will reveal the existence of any humidity 
gradient in the atmosphere, so that the effect of ventilation near doors 
and windows can be readily assessed. When daily readings are to be taken 
in a room, the instrument must be used in the same places, under the same 
conditions, and at the same time of day if results are to be comparable. 
When mildew is detected in a cupboard or among books, it is obvious 
that readings taken at the door or near a window, where the results would 
be influenced by ventilation, would not be a true indication of the atmo- 
spheric conditions in the danger area. Readings should be taken at the 
point where damp is suspected. In making a thorough assessment, con- 
ditions at night should not be neglected because in a closed room the 
relative humidity of the atmosphere increases with fall of temperature. 
The effect is most striking when heating appUances are of a type that 
become suddenly cold, such as steam radiators or electric fires. 

' National Physical Laboratory, D.S.I.R., Measurement of Humidity, 1953. 

^ Hygrometric Tables, on Ivorine Cards, Negretti & Zambra Ltd., London; 
Psychrometric Tables, Marvin, C. F., U.S. Weather Bureau, 1941. 


B 6157 



Self-recording types of hygrometer are essential when it is necessary to 
obtain a continuous record. In the usual recording hygrometer an ink line 
is drawn by a pen on a paper cylinder which is revolved by clockwork. 
The up-and-dovm movements of the pen are controlled by the contrac- 
tion and expansion of some moisture-sensitive substance, e.g. strands of 
hair, stretched within the instrument. By this means it is possible to obtain 
a record of the relative humidity as a graph on a time scale. Readings from 
self-recording hygrometers may he misleading, however, unless such 
instruments are cahbrated by a weekly check with the whirling hygro- 
meter (the more reUable instrument), and the humidity graph should be 
armotated at the time of observation. In this way it is possible to observe 
the amount of any sHght error on the chart which may arise from slow- 
ness of response due to friction, &c., and the correction is taken into 
accoimt in interpreting the records. 

Dial forms of non-recording paper hygrometer have the merit of 
compactness and are convenient for use in restricted spaces, but suffer like 
the self-recording types from a tendency towards slow response to rapid 
changes in the moisture content of the air. 


This is an accelerated decay test, and its value lies in tliat it will reveal in 
the course of a week whether a vegetable-tanned leather is suitable for 
bookbinding, i.e. whether it will survive exposure to a polluted atmo- 
sphere without deterioration. It may be apphed to all vegetable-tarmed 
leathers available today, but the test would not be appHcable in the case of 
leathers coming on the market in which the iron has been sequestrated 
(see p. 36). 

The test is carried out as follows: 

A sample of leather, 2 J in. square and weighing from 2 to 6 grams, is 
laid on a glass plate, flesh side upwards, and evenly moistened with sul- 
phuric acid (5 per cent.) in the proportion of i ml. per gram of air-dried 
leather. The acid can conveniently be apphed with a capillary pipette, 
and smoothed out on the leather with a glass rod. After rema inin g at 
room temperature overnight, hydrogen peroxide (10 vol. strength) is 
added eveHy to the leather, dropwise, in the proportion of 0’6 ml. per 
gram of leather. It is left for twenty-four hours, and then given five 
further daily doses of hydrogen peroxide; this treatment will cause un- 
satisfactory leather to be blackened and gelatinized, but durable leather 
will survive, except for possible discoloration of the edges. Changes in the 
colour of the dye-stuffs are immaterial. 


Dissolve 2 oz. of potassium lactate in i pint of water. Vegetable-tanned 
leather that is free from any sign of chemical deterioration may be pro- 
tected against the deleterious action of sulphur dioxide by spraying or 
sponging it with this solution on both sides of the skin. This ensures, as 
far as possible, that freshly tanned leather will be a durable material. 

The solution should be appHed to the outside of leather bindings that 
have been washed, in order to replace protective salts that may have been 
washed from the surface of the leather. 

Chemical decay, when once started, cannot be arrested by applying 
potassium lactate; the treatment is a form of protection, not a cure. 

Do not store potassium lactate solution for long periods of time; it 
may be preserved for a limited period in a stoppered bottle by adding 
a httle chloroform, but it is better to use the solution fresh and discard 
any residue. 

It is unnecessary to apply lactate solution to parchment or vellum, or to 
alum-tawed bindings, as these are not subject to chemical deterioration by 
sulphur dioxide. 


The ingredients are: 

LanoHn (anhydrous) 

Cedarwood Oil 

Beeswax .... 
Hexane (or Petroleum Ether B.P 
60-80° C.) .... 

7 02. or 200 g. 

I fluid 02. or 30 ml. 

I 02. or 15 g. 

II fluid 02. or 330 ml. 

These are compounded to form a yellow cream. It is highly inflammable, 
and no naked hght must be allowed in the room during apphcation of the 
preservative and for some time afterwards. It should be applied sparingly 
and rubbed into the leather, and two days later the surface may be 
polished with a soft cloth or brush. 

This dressing is obtainable from Messrs. Baird & Tadock (London) 
Ltd., 14 St. Cross Street, Hatton Garden, E.C. i. 


Silica gel is an efFecrive drying agent which has the considerable 
advantage that it does not become moist to the touch in use, nor does it 
cause staining; in fact the appearance of the material remains unchanged. 
Since it is essential to have some means of telling when it is ‘spent’ and no 
longer able to absorb moisture, it is tinted during manufacture with 
cobaltous salts which are deep blue in the dry condition, but become pink 
when sihca gel has absorbed moisture. When sdica gel has assumed a 
pink colour, it can be readily reconditioned by heating in an oven until 
it regains its original deep-blue colour. This cycle of reheating can be 
repeated indefinitely. 


There are two types, the aqueous emulsion type and the conventional 
solvent type, and the latter is generally preferable for apphcation as a 
surface coating to antiquities and works of art. 

A microcrystalhne wax salve may be apphed with the object of re- 
moving surface dirt, adjusting the optical quahty, enhancing the appear- 
ance, or excluding moisture, and when compounded with a Polythene wax 
as described m the following recipe it has been found to stabdize painted 
or varnished surfaces that are hable to bloom. 

Several varieties of microcrystalline and Polythene waxes are available 
and the gloss of the synthetic wax film can be varied by altering the grades 
and proportions of the waxes used. 

The following basic recipe has been found to be generally satisfactory: 
CosmoUoid 8o Hard^ (too g.) and BASF Wax (25 g.) are cut 
into small pieces and melted together, care being taken to ensure that 
the Polythene wax is thoroughly dispersed. The molten mixture is poured 
quickly into white spirit (300 ml.), taking precautions against fire risk, 
and, while coohng, is constantly stirred so that a paste of agreeable 
consistency is obtained. This is stored in screw-top cosmetic jars. 

Matt waxes can be produced by using CosmoUoid Soft in the pro- 
portion of 10 parts to I of Polythene wax. 

[It should be noted that although the white spirit evaporates, it may 
have a softening action on varnish of the polycyclohexanone type (AW2). 
As this is not the case with petroleum ether (boiling range 80-100° C.), 
the latter should be used as the diluent instead of white spirit where 
varnishes based on Resin AW2 are to be waxed.] 

The basic formula can be modified for use in the Tropics as a leather 
dressing and it will afford protection against the attack of insects if lauryl 
pentachlorophenate is incorporated in a concentration of about 10 per cent. 
When a wax pohsh is required for preserving bright steel, about 10 per 
cent, of sodium benzoate should be incorporated in the standard formula. 

’ CosmoUoid micxocrystallme paraffin wax, marketed by Astor, Boisellier,& Laurence 
Ltd., 9 Savoy Street, London, W.C. 2. 

^ BASF Wax A (Polythene wax), marketed by Bush, Beach, & Gent Ltd., Marlow 
House, Lloyd’s Avenue, London, E.C. 3, is now obtainable in the form of peUets. 


Apparatus Required: 

One Transformer i8 v. 12 a. 

One Bridge-Connected Funnel-Cooled 
Selenium Rectifier 17 v. 12 a. 

One Variable Resistance 3-5 ohm 12 amp. 

One Ammeter 0-15 a. M.C. Flush Mount. 
One Pair of Large Terminals; One Red, 
One Black (Belling Lee). 

One Pair of Slydlock (or other) Fuses. 


Abrasives: for polishing metals, 205-6, 213, 
216, 231, 257, 268, 291; for rubbing 
down stonework, 313-14. 

Acclimatization of waterlogged objects, 


ACP Deoxidine No. 125 , 277, n. 3. 

Adhesives, used in repairing: baked clay, 
324; bronze, 255; canvas p ainting s, 167- 
8, 169; clay tablets, 323; firail textiles, 
106, 108; glass, 339; ivory and bone, 152; 
leather, 32; marble, 313, 317; painted 
panels, 160, 164; pottery, 329-32; seals, 
91-92; steel, 285; stone objects, 317, 318- 
19; tears in paper, 86, 87, 347-8; wood, 

Aerodux No. 185 , 330-1. 

Aerohte No. 300 , 132. 

Agora, lead objects from, 261. 

’Ain Feshkha, Palestine, finds at, 40. 

Air conditio ning , 4 ff. & n. 

Alabaster, 295. 

Alcho-Re solder paste, 288. 

Alcohol-ether-resin, and waterlogged 
wood, 136-8. 

Alginate, use for moulding, 153. 

Alloys, 188-9, 189 n.; decay and, 190; 
drying of, 199; cleaning of, 205; of gold, 
207, 208; of silver, 220, 221, 224, 230; 
of copper, 232 ff., 248, 256; gun-metal 
and bell-metal, 247; speculum, 248-9, 
288; of lead and tin, 258, 270; and 
soldering, 286. 

Alum, 23, 24, 25, 52, I35-<5, 139, 347- 

Amber, 156. 

Amberhte IR No. 120 , 263 n. 

Antimony, 258. 

Antisun glass, 1 12 n. 

Antler, 155-6. 

Antrobus, Mrs. Guy, 109. 

Araldite No. 101 , 168 & n. 2, 330, 
339 - 

Armitage, F. D., 52 n. 2, 54 n. 2. 

Athenaeum Club, 19. 

Atkinson, R. J. C., 327 n. i. 

Atmospheric pollution, 4-5, ii ff., 213, 259, 
277. 316; dust, 95, 171, 210, 307; paint 
finnes, 259; sea-air, ii; soot, 4, ii, 95, 
171, 210, 297, 307; sulphurous gases, ll- 
12, 19, 34 , 35 , 36, 37 , 52-53, 112, 166, 
170-1, 188, 213-15, 277. 

Atomizers, 30, 115. 

Australia, Hartogs plate from, 266. 

Bacteria: and leather degradation, 22, 26; 
and textiles, 93 ; sulphate-reducing, 
138-9, 272-3, 282; and water paints, 213 ; 
and iron corrosion, 272-3. 

Bakelite, 130 & n. 3. 

Baker’s Soldering Fluid, 287. 

Bark-cloth, 133-4. 

Barker, H., 244 n. 

Barrow, W. J., 53, 61 & n. 2, 62, 
65 n. 

Basalt, 296, 297. 

BASF Wax A, 359 & n. 2. 

Basketry, 133, 

Bastille medal, 268. 

Battiscombe, C. F., 108 n. i. 

Bedacryl 122 X, 130 & n. i, 257, 304. 

— L, 151. 

Beetles, wood-boring, 121-3. 

Belgian Museums, Central Laboratory of, 
Brussels, 128 & n. 2. 

Bellinger, Louisa, 105. 

Bengtson, Bengt, 215. 

Bennister, H. L., 279 n. 2. 

Biek, L., 279 nn. 2 & 3. 

Binding mediums in painting, 158, 169- 

Birch bark, 44. 

Black Prince relics, 107 & n. 

Bleaching: of paper, 59-60, 64-65, 75-76; 
of stained prints, 74 ff. ; of old textiles, 
100; of waterlogged wood, 136, 

— agents, 75 ff. 

Blistering, paint, 164-6. 

Bloom, 10, 172-3. 



Blotting-paper, use of, for: relaxmg 
papyrus, 43; print cleaning, 70 ff., 83; 
crease-removal, 85; mounting, 88; ab- 
sorbing grease stains, loi, 108; applying 
reagents, 148, 149, 154; flattening 
wooden panels, 163. 

Boats, recovery of, 140-3. 

Bone, composition of, 144-5; preservation 
and restoration of, 145 ff. 

— objects, 2, 144, 146; conservation of, 

147-8, 151- 

Bookbinder’s paste, 54, 347. 

Bookbindings, 9, 10, 12, 19, 24, 25-26; in- 
secticides and, 30; deterioration of, 34- 
35; ‘peroxide’ test and, 35-36, 37. 355: 
preservation of, 36-39; ivory, 144. 

Book of the Dead, 42. 

Books, 9; mould growths on, 10, 27, 55; 
sea air and, ii; coal gas and, 19; binding 
of, 25-26; attacks by insects on, 30; 
causes of deterioration, 38; dressings for, 
38; washing of, 38-39; use of parchment 
for, 48; paper for, 52; sterilization of, 57, 
58-60; repair of, 61, 72, 88; removal of 
pages from, 84; ivory covers for, 

Bradley, M. C., 178 n. 2. 

Brass, 189, 283 n. 5. 

British Honduras, lava monuments from, 

— Leather Manufacturers’ Research As- 
sociation, 19 & n. 2, 35 & n., 37 n. 

— Museum, 4, 54, 123, 147; objects referred 
to: leather bindings, 34-35, 36-37; King’s 
Library, 46; Sutton Hoo finds, 116; clay 
tablets, 320; Chinese frescoes, 323. 

Leather Dressing, 38, 357. 

Research Laboratory, 3, 60, 105, 138, 

153, 208, 225, 262, 301, 321, 337, 338. 

— Records Association, 58. 

Brommelle, N., 172 n., 176 n. 2. 

Bronze, 139, 188-9, 233; decay and, 190; 

‘disease’, 234-6, 246, 252, 256,272,273; 
decorated with tin, 265. 

— Age enamel, 338. 

— objects; treatment of incrustation on, 
192, 193, 197, 198, 204, 205, 210, 234 ff ; 
reconstmction of 286. 

Bronzes, Chinese, 94, 188, 247-9, 250, 251. 

Brushing, of metaUic objects, 192, 196, 
197, 203-4, 221, 222, 224, 227, 253; and 
the appHcation of wax, 280-1 ; of stone 
objects, 298; of marble, 309. 

Buddhist documents, 44. 

Burgess, S. G., 273 n. 2, 296 n. 

Burial, effect of, on: alloys of gold, 207; 
bone and ivory, 145; clay objects, 320; 
copper, 233, 254; iron and steel, 272-3, 
278; metals, 187-8; silver coins, 224; tin 
and lead, 258, 259, 265. 

Burnishing, 205-6. 

Budin, K. R., 273 n. i. 

Caley, E. R., 197 & n., 261 & n. 

Calgon, 247. 

CaUigraphy, 65. 

Canoes, recovery of, 140-3. 

Canova, 311. 

Canvas, see Paintings, easel. 

Carbon, in ink, 63-65, 328; and steel, 189. 

— disulphide, 124-5, 346- 

Camehan, 153, 154, 294. 

Carpets, 98-99. 

Carrara marble, 308. 

Carruthers, R. H., 58 n. i. 

Carson, F. T., 53 n. 3. 

Cartouches, 253. 

Carving, ivory and bone, 144, 152-4. 

Casein, 311. 

Catechol group of vegetable tannins, 25, 37. 

Cathodic protection, 189, 219, 250, 259, 

Caustic soda: and metalhc reduction, 191 ff., 
259 ff. , 276, 278; care when using, 345, 


Cellophane, 29. 

Cellulose acetate, 61-62, 133. 

Cements, for repairing: stone objects, 318- 
339 - 

Cennini, Cennino, 158 n. i. 

Ceramics, see pottery. 

Chakravorti, S. K., 59 n. i. 

Chalk, 158, 193. 

— drawings, cleaning of 83, 84. 

Chatterton’s Compound, 132. 

Chemicals, dangerous, 345-6. 

Cheshire, A., 36 n. i. 


Chinese: brass, 189; bronzes, 94, 188, 247- 
9, 250, 251; frescoes, 323; inks, 63; 
ivories, 151; lacquer, 2; paper-making, 
50-51; pottery, 329; silk, 93. 

Chipping metallic incrustation, 202-3. 

Chloramine-T, 77, 79, 83, 308, 309. 

Chloride, test for, 198, 199, 237-8, 241, 
243, 298; and rust, 272, 273-4. 

Chlorine dioxide, 60, 75, 77. 

Christensen, B. B., 137 and n., 138. 

Ciba Experimental Dye-house Labora- 
tories 99. 

Clay, 205, 273, 319 ff.; for writing upon, 
320; methods of firing, 326; see also 

— objects, 319; treatment of, 320 ff. 

Cleavage in easel paintings, 158,159, 164, i6d. 

Cleopatra’s Needle, 296. 

Cockerell, D., 86 & n. 

Cody’s Tree, 130. 

Coins: embedding of, 267-8 ; electrum, 207; 
gold, 207, 209, 212; lead, 264; Roman, 
189; silver, 212, 224-5; Sutton Hoo, 26; 
tin, 265, 266. 

Coleoptera, 28, 121. 

Consohdation, see reinforcement. 

Constable, W. G., 181. 

Cooper, B. S., iii n. i. 

Copper, 188-9, 195. I 99 > 267; in gold 
alloys, 207, 208; gilding of, 209; in- 
crusted on silver, 219-22, 223-4, 231; 
oxidation of, 232-4, 239; conservation 
of, 234, 249; treatment of, 239 ff., 279; 
decoration of, 249 ff.,; alloys of, 256; in 
pewter, 258; and soldering, 289. 

Coptic stele, 301-2. 

Coral, 156. 

Coraline solder paste, 288. 

Cored Ersin Solder, 288. 

Coremans, P., 128 n. 2. 

Corrosion, 93, 138, 139, 185; treatment of, 
186, 191 ff.; in buried metals, 187-8, 
248; process of, 189-91, 271; mechanical 
removal of, 200 ff.; of silver coins, 224- 
5; of copper and bronze, 233 ff.; of lead 
and tin, 258-9, 265-6; of pewter, 268; 
of iron, 271-4; soldering and, 287. 

Cosmolloid microcrystalline waxes, 359 & 
n. I. 


Cracking of wooden panels, 159, 160, 161, 
163, 166. 

Craquelure, 141, 142, 158-9, 171. 

Crepehne, 61. 

Cripps, E. S., 279 n. 3. 

Croffles, 196. 

Cupping of painted panels, 118, 163. 

Curtains, cleaning, 98-99. 

Damascene work, 281. 

Damask, cleaning, 98. 

Damp: staining by, 5, 21, 47, 55; degrada- 
tion of leather by, 26, 3 1 ; and insect pests, 
58; and textiles, 93-94, 114; effect of, 
on wood, 116-17, 120; effect of, on bone 
and ivory, 144-6; and metalhc oxida- 
tion, 188, 233-4, 245, 274; and stone 
waxing, 303 ; and marble, 307-8. 

Da Vinci, Leonardo, 66. 

Dayhght, see sunlight. 

DDT, 29, 59, 1 15, 125. 

Dead Sea Scrolls, 40. 

De Bruyne, N. A., 132 n. i. 

Decay, electro-chemical, 189-90. 

Decaying matter, and textiles, 93, 95. 

Decoration, concealed, 14; cleaning of, 104, 
299. 327, 328; with ivory, 144, 153; 
with gold, 209-10, 249-50; with silver 
and tin, 249, 250 ff. 

Delougaz, P., 320 n. 

Delta metal, 313, 318. 

Desch, H. E., 118 & n. 2. 

Desiccation, damage by, 9, 21; and skin 
products, 31; by siUca gel, 31, 200, 245, 
246, 267, 336, 358; and parchment, 47; 
its effect on wood, 116, 118-19, 132-3, 
161; of metals, 199-200, 223, 246. 

Detergents, for use with: bone and ivory, 
147, 154; glass, 335; leather, 33 n.; 
marble, 307, 316; silver, 229; textiles, 97. 

Devitrification, 329, 335. 

Diastase, 105 & n. 2. 

Dorchester, silver coins from, 225. 

Dowelling, 160, 317-18, 329. 

Drawings: sterilization of, 57; bleaching of, 
60, 75-81; study of, by photography, 66; 
cleaning and repair of, 68 ff.; mounting 
of, 88-91. 

Dry-cleaning, of prints &c., 72-73; of 



textiles, 98-ioo;ofcorrocled metals, 204- 
5, 252. 

Dry rot, 120 & n. 3. 

Duerden, H., 310 & n. 

Dura-gHt, 216, 227, 257. 

Durham vestments, 107-10. 

Durofix, 140, 152, 154, 228, 255, 266, 285, 
323, 329. 339 - 

Dusting of marble, 306-7. 

Dye-stuffs: soap and, 33 n. ; and leather, 35; 
desiccation and, 47; in inks, 65; removal 
of. 75. 79; hi cleaning, 81, 82; mould 
growths and, 83-84; in textiles, 95, 96, 97, 
98, 108; fixing of, 96, 97; organic sol- 
vents and, 99, 100; and Terydene, 106-7; 
hght and, iii; and ivory, 146-7. 

Edwards of Halifax, 46. 

Egg-tempera, 158, 169, 170, 175. 

Egypt, ancient: skin products in, 21; tan- 
ning in, 25; tomb paintings from, 25, 
299; leather scrolls from, 39-40; use of 
‘paper reed’ (papyrus) in, 41-42; writing 
fluids in, 63 ; mummy wrappings from, 
94; wood from tombs of, 116; ivory' 
objects from, 144, 148, 149-50; bronzes 
from, 193, 244, 246-7, 252-4; gold ob- 
jects from, 208; nieUo decoration in, 221 ; 
and alabaster, 295; vessels from, 327; 
glass scarabs, 336-7. 

Electrolytes, and corrosion, 189-91; reduc- 
tion by, 194-7, 222, 223. 

Electro-plating, 289. 

Electrum, 207 ff. 

El Greco, 170, 

Embedding in plastic materials, 267-8. 

Embroidery, 14, 93 ; cleaning and mount- 
ing of, 107-10. 

Emesa Helmet, restoration of, 226-9, 288. 

Enamel, 337-9. 

Enkomi Cup, 220. 

Engravings: bleaching of, 60, 75-81; 
cleaning and repair of, 68 ff.; mounting 
of, 88-91. 

Environment: influence of, i ff.; and 
identification of objeas, 13; and painted 
panels, 160-1; and canvas paintings, 166; 
and corroded bronze, 235. 

Enzymes, 344. 

Ercaline lacquer, 229, 244, 253, 257. 

Etchings; bleaching of, 60, 75-81; cleaning 
and repair of, 68 ff; mounting of, 88-91. 

Ethnographical material, 10, 20, 27, 100, 
115, 116, 121, 323; sterilization of, 123. 

Excavated objects, i, 2-3 ; bone and ivory, 
151; bronze, 235, 251-2; clay, 320, 327; 
lead, 259; silver, 216-17; textile, 103. 

Faraday, Michael, 19 & n. i. 

Farnsworth, Miss M., 247 n. 

Fibres: animal and vegetable, 93, 117; sun- 
fight and, no; synthetic, 107. 

Fillers: for repairing — wood, 132; ivory, 
150; silver, 228; by dowelling, 318, 319. 

Fire, precautions against, 50; deciphering of 
writing after, 67; effect of, on bone and 
ivory, 145, 146; gilding by, 209; effect 
of, on marble, 312, 313. 

Fisher, R. C., 122 n. 

Flags, mounting of, 106-7. 

Flaking paint, 164-5. 

Flints, 156. 

Fluxes, soldering, 287-8. 

Fly-marks on papers, 73, 78. 

Forest Products Research Laboratory, 120 
& n. 3, 122 n., 125 n. 2, 126 & n. 

Formalin, 57, 60. 

Fossflization, of bone and ivory, 145, 148. 

Fossils; ‘pyritized’, treatment of 295; im- 
prints of 309. 

Fowler, J. N., 102. 

Foxing of papers, 53, 54, 60, 82. 

Franks Casket, 147. 

French pohsh, 331, 332. 

Frescoes, 303, 323. 

Frigilene lacquer, 229. 

Frost, effects of, 296-7, 304, 316. 

Fryolux solder paste, 288. 

Fuels, and air pollution, 12, 19, 37, 112. 

Fumigation, 28-29, 30, 49. 115; and mould 
growths, 55 ff.; and insect pests, 123-5, 

Fungicides, 27-28, 49, 56, 58, 120, 121, 181. 

Fungoid growths, 10; and skin products, 27, 
49; and paper, 54; and textiles, 113; and 
wood, 116, 118, 120, 121; and canvas 
paintings, 180-1 ; and marble, 307-8; and 
enamel, 338. 



Furniture, 9, 98, 119, 121-2, 133, 294. 

Gammexane, 59, 125. 

Gelva (polyvinyl acetate), 13 1. 

General Catalogue of Printed Books, binding 
of, 35. 

Gesso, 4, 21, 47, 107, 121, 127, 157, 166, 

Gettens, R. J., 59 & n. 3, 233 n. 

Gilding, 89, 209-10, 249-50. 

Glanville, Prof. S. R. K., 40 & m 

Glass, 333 IF.; for museum windows, 112 
& n.; composition and properties of, 
333-5; decomposition of, 335-7; repair 
of, 339 - 

— objects, 223. 

Glaze, 328-9. 

Glue, 9, 10, 22, 38, 63, 13 1, 132, 136, 165, 
167, 169. 

Glycerine, 135, 141, 142. 

Gogstad ship, 143. 

Gold, II, 188, 189, 212, 337; treatment of, 
153, 206, 207, 209-10, 250; alloys of, 
207-9; as decoration, 209-10, 249 ff. 

— amalgam, 249. 

Gold-beater’s skin, 44. 

Gold leaf, 21, 26, 107, 127, 154, 155, 158, 
246, 249. 

Gold objects, restoration of, 210-11. 

Gold thread, 95; cleaning of, 107-10. 

Gouache, 170. 

Grain, in wood, 117, 118. 

Granite, 296-7, 306. 

Greek bronze, 245; marbles, 311, 315, 316. 

Greene, F. S., 104 n., 105 n. 

Grimes, W. F., 3 12. 

Grinding, 203. 

Harrison, L. S., non. 

Hartogs plate, 266. 

Heat, staining by, 21; effect of, on textiles, 
94; effect of, on clay, 319, 320-1. 

Heating, 5, 8, 9; and leather deterioration, 

Hedvall, J. A., 335 n. 

Henshall, A. S., 94 n. 

Herculaneum, 42. 

Hermetic sealing, 12-13. 

Holden, J. T., 102 n. 

Horn, 155. 

Horridge, G. A., 130. 

Houdin, bust of Voltaire, 308. 

Houwink, R., 132 n. i. 

Hughes, A. W. McKenny, 58&n. i, 115 n. 

Humber, the. Iron Age boats from, 141. 

Humidity: changes and effects of, i if, 
27, 117, 158, 164; on skin products, 26, 
27. 30-31; parchment, 47, 49; paper 
storage, 53-54; textiles, no, in; sea- 
soning of wood, 118, 129; easel paintings, 
158 n. 2, 159; painted panels, 163, 164; 
canvas paintings, 166; rusting, 277, 
stone monuments, 296; measurement of, 

Hunter, Dard, 51 & n. i. 

Huntington Library, San Marino, 58. 

Hydrogen peroxide test, 35-36, 37, 355. 

Hygrometers, 353-4. 

Hygroscopic materials, 8, 10, 47, 49, 53, 
117. 335. 336. 

Hyphae, 10. 

Igepal CA Extra, detergent, 97, 105. 

liams, T. M., 58 n. 2. 

Illumination, in MSS., 44, 45, 47-48. 

Illumination, ultra-violet and infra-red, 
66-67, 176; examination of marble by, 

impasto, 169, 171, 174, 177. 

Impregnation, 56, 57-58, 123,125, 126, 127- 
31; methods used for: horn and antler, 
8cc., 155; incrusted ivory, 150; lead and 
tin alloys, 270; marbles, 311; pottery, 
328; ‘pyritized’ fossils, 295; silver coins, 
224; stone objects, 302-6; unbaked clay, 
323; water-logged wood, 135-8. 

Incrustation, 13; as evidence, 13-14, 188, 
203, 249, 284; of ivory, 148-50, 153; of 
metalhc objects, 185 ff., 206; of silver, 
216 ff; of copper, 233 ff; of lead & tin, 
258 ff; of pewter, 268-9; of iron, 273 ff; 
of minerals, 295, 298, 299 ff; of glass, 
336, 337- 

— removal of, by: brushing, 203-4, 216, 
222, 224, 227, 240, 241, 253; chemicals, 
207, 210, 219 ff, 238 ff, 250, 261 ff; 
electro-chemical processes, 191-7, 214, 
217, 221 ff , 235, 237, 239, 241, 253, 265, 



269, 270, 274, 291; mechanical means, 
200-5, 210, 216, 219, 231, 248, 250, 251, 
257, 264, 270, 275, 291; washii^, 197- 
9, 222, 237, 241, 244, 253, 257, 260-1, 

270, 291. 

India proof papers, 78; cleaning of, 82-83. 

Indicators, 260, 350-1. 

Inhibitors, rust, 277-9. 

Inks, 41, 42, 47, 63 ff.; puncturing by, 53; 
preserving of, 53; carbon, 63-65, 328; 
iron, 64-65; cleaning and, 70, 75, 76, 
86; stain removal, 79. 

Inlay, 9. 132, 275, 294; marcasite, 295; 
niello, 221; red glass, 338; silver, 226. 

Innes, R. Faraday, 35, 36 n. 2, 37n., 53. 

Inpainting, 163, 167, 175-6. 

Inscriptions, treatment of, 224, 253 , 266, 269. 

Insecticides, 28-30, 114, 115 & n., 121-2, 
124, 125, 126, 180. 

Insects, attacks by, on: basketry and bark 
cloth, 133-4; panel paintings, 180; parch- 
ment and paper, 58; skin products, 28 ff.; 
storage materials, 50; textiles, 1 13-15; 
wooden objects, 121-6; control of, 28- 
30, 114. 

International Institute for the Conservation 
of Museum Objects, in. 

Ion-exchange principles, 262-4, 268. 

Iron, 188, 190, 195-6, 203, 205, 271; 
in gold alloys, 208; in juxtaposition to 
silver, 226; to brass283 No. 8; rusting of, 
271-3 ; sulphides of, 295. 

— objects, 199, 271; examination of, 273-4; 
treatment of, 274-7, 282-3, 284, 290-1; 
protection of, 277-80; preservation of 
rust on, 281; reconstruction of, 284-6. 

Islamic bronze, 242. 

Istituto di Patologia del Libro, Rome, 67. 

Italian marbles, 315. 

Ivory: composition of, 144-5; preserva- 
tion and restoration of, 145-55. 

— objects, 2, 144, 146; conservation of, 
147-55; moulding of, 152-3. 

Jade, 294. 

Japanese: ivories, 146; mulberry tissue, 
108; prints, 82, 87; swords, 206, 280, 283 ; 
vellum, 82. 

Jet, 156. 

Jewellery, 156, 205, 207, 265. 

Jones, G. A., 67 n. i. 

Kantrowitz, M. S., 51 n. 2. 

Keeley, T. R., 4 n. 

Kimberley, A. E., 53 & n. 2. 

Kramer, W., 136, 138. 

Lace washing, 95, 96. 

Lacquer objects, 2-3. 

Lacquers, used in protecting: clay objects, 
323, 329; copper and its alloys, 244, 245, 
253, 257; enamel, 338; fossils, 295; glass, 
335 & n.; iron and steel, 279; lead, 
265; silver, 215, 229, 231; stone objects, 
303 - 5 - 

Lamination, of paper, 60-63, 7°. 89. 

LangweU, W. H., 53 & n. 

Lanolin, 277. 

Lapis lazuli, 153, 154-5, an, 294. 

Lassoband adhesive tapes, 154 &n. 

Lavas, 305. 

Lead, 188-9, I 97 > 259; in Chinese bronze, 
248; in solder, 286-7; hi glass, 333 & 
n. I, 337 - 

Leaden objects: treatment of incrustation 
on, 198, 258 ff.; oxidation of, 258-9; 
washing of, 260-1; acid treatment of, 
261-2; ion-exchange resins and, 262-4; 
storage of, 264-5, 270. 

Leather, 9, 10, 19, 55 n. 1; dressing of, 20- 
21, 32-33, 37-38; preservation and pro- 
tection of, 21-26, 30-32, 35-39. 356, 357; 
for bookbinding, 25-26,34^,355; deteri- 
oration of, 26-28, 34-35; fungicides for, 
27-28; moth attacks on, 28; repairing of, 
32-33; vegetable-tanned, 34-39, 356; 
peroxide test for, 355. 

Libraries, mould growths in, 10, 27, 58; 
care of books in, 38, 

Library of Congress, Washington, 13 n. 

Lighting: in museums, 111-12; fluores- 
cent and tungsten, 111-12. 

Lime, 193, 207; deposits of, on bronze, 247. 

Limestones, 298, 300, 301, 305. 

Liquid Ersin soldering flux, 288. 

Lissapol detergent, 60, 74, 97, 103 n., 105, 
133. 207, 221, 229, 297, 307. 

Lucas, A., 25 & n., 41-42, 176 n. 2. 

Lumarith, 10411. 

Lumite, 104 n. 

McBumey, C. B. M., 151 & n. 5. 

McIntyre, J., 8 n. i, 54 n. i. 

Malachite, 294. 

Malaya, coins from, 266; kris from, 281. 

MaUowan, Prof. Max, 153. 

Manuscripts: mould growths on, 55; steri- 
lization of, 56-57, 5 8-60 ; repair of, 61, 68 ; 
mounting of, 89-92; use of gold in, 

Maps, mounting of, 63. 

Marble, 161, 306 ff. ; dusting and washing 
of, 306-7, 309, 316; stains on, and their 
removal, 307-10, 313-14; warping of, 
310; granulation of, 310-11, 313; patina 
of, 314-15; examination of, by ultra- 
violet hght, 315-16; effect of weathering 
on, 316-17; packing of, 317. 

Marcasite, 295. 

Marco Resin S.B. 26C, 130 & n. 2 267- 
8 . 

Marquetry, 9, 133. 

Martin, R. S, J., 151 & n. 4, 267 & n. 2. 

Maryon, H., 201 n. 2, 211 n., 283 n. 3, 285 
& n., 286 n. 

Matson, F. R., 334 & n. 

Maxwell, S., 94 n. 

Mayer, R., 181. 

Meeuse, Dr. A. D. J., 115 n. 

Mercury gilding, 209, 250. 

Merovingian objects, 13 n., 283. 

Metalhc objects, 3, ii, 14, 26, 161, 295, 
321; conservation of, 185 ff., 192 ff.; 
cathodic protection of, 189-90, 219; 
reduction methods and, 191 ff., 235; 
drying of, 199-200; mechanical opera- 
tions and, 200-3; pohshing of, 205-6; 
gold decorations on, 209-10. 

Metals: characteristics of, 185 ff.; effect of 
burial on, 187-8; of antiquity, 188-9, 
234-5, 284, 286-9; cathodic protection 
of, 189-90, 250, 259; incrustation and, 
198; drying of, 199; gilt, cleaning of, 
210; soldering of, 286-9; patina of, 314; 
used for doweUing, 317; as a base for 
enamels, 337. 

Meteorites, 271. 

INDEX 367 

Metropohtan Museum of Art, N.Y., 255. 
Michael Angelo, 308. 

MicrocrystaUine wax, 33, 257, 280, 303, 
359 - 

Micro-organisms: and skin products, 20, 
46; and paper, 53-54; and wood, 120. 
Micro-structure, as evidence, 14; of ‘skin’, 
21-22; of wood, 116, 118; of bone and 
ivory, 145; of gold alloys, 208; of silver, 


Mildew, see mould growths. 

Mills, J. S., 176 n. I. 

Minerals, 294-5. 

Miniatures, painted, removal of stains 
from, 81. 

Mithras Head, treatment of, 312-13. 
Moisture, in wood, 117-18. 

Monuments, stone, 294, 296; preservation 
of, 296-7. 

Moss, A. A., 37 n., 136 &n. i., 221 &n. 2, 
275 n., 283 n. I. 

Moss, A. J. E., 100 n., 102 n. 

Moth: control of, 28-29, 30 . 114-15, 
115 n.; and textiles, 99, 114. 

Mother of pearl, 156. 

Mould growths, 5, 9-10, 22, 26, 43, 353; 
and skin products, 10, 27, 31, 46, 49; 
removal of, 31, 72-73; on paper, 54, 82; 
methods of dealing with, 54-58; on 
pastels, 83-84; on chalk drawings, 84; 
and textiles, 93, 95, 106, 113 ; and wooden 
objects, 121; and basketry, 133. 
Moulding, of ivory, 152-3. 

Moimting, 88-92, 106-7, 108-9. 

Mulberry fibre, and paper-making, 50- 


— tissue, 164. 

Midticore, ARAX soldering flux, 288. 

— solder paste, 288. 

Miuray, H. D., 67 n. 2. 

Museums, climate of, i, 8, ii, 13; causes of 

damage in, 15; and moth control, 28; 
lighting of, 111-112; sterilization of, 
123 ff. 

National Bureau of Standards, U.S.A., 
13 n., 27 & n., 52 n. i, 53 nn. 2 & 3. 

— Gallery, London: air-conditioning in, 
4 & n., varnish used by, 176. 


Paper, 9, 10, ii, 12, 41, 44; advent of, 42, 


National Library of ^J^ales, Aberystwyth, 


— Maritime Museum, 141. 

— Museum of Antiquities of Scotland, 
Edinburgh, 94 n. 

Neri, A., 333 &n. 2, 334. 

Newspapers, 52, 53. 

Niello, 220-2. 

Nimrud, ivories, 146, 153-5; cuprous 
enamel, 338. 

Nital, 280. 

Nitro-cellulose treatment, 299-300, 309- 
10, 328. 

Noric Museum, Stockholm, 215. 

Nonex detergent, 74. 

‘Non-tans’, 35, 37. 

North Ferriby, 141. 

Nylon bags, use of, 195, 196. 

Old Masters, 168, 170, 172. 

Optical quality, 129, 157, iji, 172, 179. 
Organ, R. M., 198 & n., 199, 244 n., 262 n., 
266 n. 

Orgamc matter, staining by, 310, 314. 
Orichalcum, 283 n. 5. 

Ornamentation: of metallic objects, 186, 
192, 216, 220, 241, 242, 244, 266, 275; 
niello, 220-2; reconstruction of, 285. 
Oseberg ship, 143. 

Oslo, Viking ships at, 142-3. 

Ostraka, 328. 

P 84 sodium sihcate, 141. 

Packing, of marble, 307-8, 317. 

Paimings, effects of humidity on, 4, 9, 11; 
effects of atmospheric pollution on, ii; 
ancient Egyptian tomb, 25, 299-300,’ 
see also Alimatures. 

Paintings, easel: structure of, 157-9- re- 
i^orcement of, 161-3; repair of, 167-9; 
cleaning of, 169-80; canvas, 165-81- 
wooden panel, 119, 124, 1270., 160-5’ 
metal panel 161. 

— tempera, 158 & n. i, 171, 175. 

— water:<olour, 170; sterilization of, 57- 

bleaching of 77. ’ 

Paints: composition of ij8, 169-70; causes 
ot detenoration in, 170-r; and silver 
tarnish, 213; staining by, 309. 

50-51; raw materials for making, 51- 
52 , tests for permanence of, 52 ; acidity of 
52-53, 62; storage of 53-54; sterilization 
of against mould, 5 5-5 8 ; insect pests and, 
58-59; resizing and bleaching of 59-60; 
lamination of, 60-63, 70; examination of 
before cleaning, 69; removal of creases 
from, 85-86; repair of tears in, 86-87; 
resizing and toning of 87-88; mounting 
of 89. ® 

pulp. 246; and the treatment of stone 
objects, 299-300, 313, 328. 

Papyrus, 39, 48; making of 41-42; from 
^SyP’uan tombs, 42-43 ; method of 
unwrapping, 43-44; sterilization of 
58, ink used on, 63; bleaching of 
75 - 

Paraffin oil, 277. 

— wax, 32, 179, 297. 

Parazyme digestive enzyme mixture, 105 
Parchment, 9, 10, 21, 24, 40, 41, 55 n. i; 

44-45; manufacture of 45-46; 
alkalinity of, 46-47; humidity and, 47- 
48, 49; relaxing of 48-49; storage of 
49-50; sterilization of 57-58; insect 
pests and, 58-59; removal of creases 
from, 85; repair of tears in, 87; mount- 
ing of 88. 

Parthenon, the, 31 1, 316. 

Partington, J. R., 212 & n. 

Paste, flour, preparation of, 347-8. 

Pastels, sterilization of 57; cleaning of 
83-84; mounting of 89. 

Patching, of canvas, 167; of ancient metals, 
289; of reconstructed pots, 331 
Patina, 187, 188, 205, 235; on bronze, 235, 
236, 238, 239, 2428".; gold, 209; lead, 
258-9: marble, 314-15; pewter, 268; 
silver, 215-16; stones, 297. 

Patterns in sword blades, restoration of, 

Pearson’s Square, 349. 

Pease, M., 255. 

Peat, objects excavated from, 94 & n., 116 

Pergamon, Asia Minor, 21, 44. 

Peroxide (or ‘P.I.R.A.’) test, 35-36, 37, 



Persian: bronzes, 243; kermes, 65; minia- 
tures, 81; ‘parcel gUt’ decoration, 209; 
pottery, 329. 

Perspex, 13 n., 105, 254-5, 268, 336, 339. 
Petrie, Sir Flinders, 42. 

Pewter, 258; cleaning of, 268-9. 
Photography: and reading faded writing, 
66-67; Mid the repair of textiles, 104; 
and detecting restoration, 176. 

Picking, 201-2, 203, 241, 250, 264. 
Pigments, 158, 169, 174,i76;tintingplaster 
for repairing pottery, 331-2. 

Plaster: on bronze, 246; casts, 286. 

— of Paris, 318-19, 327, 330, 331-2. 
Plasticine, 253, 331. 

Plastics, for embedding coins, 267. 

Plesters, Joyce, 267 & n. 3. 

Plenderleith, H. J., 37 n., 54 n. 3, 59 n. 2, 
150 n., 151 n. I, 266 n. 

Plus-Gas Fluid A rust softener, 277 & n. i. 
Polishing of metals, 205-6. 

Polythene, 29, 73, 95-96, 104, 114, 124, 
163, 196, 214, 325, 336, 339- 

— wax, 179, 359. 

Potassium lactate treatment, 35, 39, 356. 
Pottery, 326 ff.; cle anin g of, 327-9; repair 
of, 329-32- 

Prince Rupert drops, 337. 

Prints, sterilization of, 57; bleaching of, 60, 
64-65, 75-81; cleaning and repair of, 
68 ff.; mounting of, 88-91; see also 

Purves, P. E., 151 & n. 4, 267 & n. 2. 
Pyrogallol group of vegetable tannins, 25, 


Pyx Cloth, 107. 

Quartz, 207. 

Radiation, ultra-violet, 176. 

Ragging for local electroplating, 228, 289. 
Rags, and paper-making, 50-51. 

Rawlins, F. I. G. 4 n. 

Reagents, for removing stains, 102-3. 
Reduction, chemical and electro-chemical, 
ofmetalhc incrustation, 186, 191-7, 217, 
221, 231, 237-8, 239, 253, 257, 259, 269, 
270, 274, 275, 291, 360. 

Refrigeration, 40. 

Reinforcement, methods used with: clay 

B 8167 B 

tablets, 323; copper alloys, 245, 257; 
enamel, 3 3 7-8 ; marbles, 3 1 1- 1 2 ; oxidized 
tin, 266; pottery, 328, 329; rusted iron, 
281; silver, 227-8, 231; stone objects, 
299-300, 301-6. 

Relining easel paintings, 166, 168-9. 

Resin AW 2 and MS 2, 179 n. 

— varnishes, 176-9. 

Resins, Epoxy, 13 1, 168. 

— used for: canvas repairs, 167, 168, 169; 
consolidating wood, 129-30; embedding 
coins, 267-8; flaking paint, 164; fluxes, 
287; ion-exchange, 262-4; ivory repair, 
150; pottery repair, 330; wood repair, 
131, 136. 

Restrainers in controlling solvent activity, 

Rhind Mathematical Papyrus, 39. 

Rijksmuseum, Amsterdam, 266. 

Rochelle salt process, 210, 239, 224, 225, 
239-42, 250, 251, 252. 

Rock crystal, 294. 

Rocks: igneous, 296-7, 306; sedimentary, 

Rodents, attacks by, on parchment, 50. 

Rolls, leather, unrolling of, 39-40. 

— , papyrus and parchment, sterilization 
of 58. 

Roman scabbard of brass, 283 n. 5. 

Romano-British steelyard, 282. 

Rosen, D., 128 & n. i. 

Rosenberg, G. A., 135 & n. 

Royal Academy of Arts, 308. 

— Institution, 19 & n. i. 

— School of Needlework, 107 n. 

— Society of Arts, 19. 

Rubber-paste cement, 167. 

Ruhemann, H., 169 n. 

Rust: removal of 200, 202, 204, 205, 206, 
227,274-8, 291 ; production of, 271-2 ; ex- 
amination of, 274,284; inhibitors, 277-9; 
preservation of, 281, 291 ; as evidence, 284. 

St. Cuthbert rehes, 108, 152. 

Salty ground, contamination in, 3, 145, 
148, 190, 216, 233, 259, 320, 327, 328. 

Sand, 1 16, 207. 

— blasting, 321. 

Sandstones, 298, 300, 305, 306. 




Santobrite, 28, 58, 84, 181, 338. 

Santucci, L., 67 n. 4. 

Saponins, glucosides used as detergent, 98, 
108, 109. 

Saxon bronzes, 251, 254, 338. 

SebaefFer, C. F. A., 220 n. 

Schaffer, R. J., 296 n. 

Scraping, 202-3. 

Scribner, B. W., 52 n. i, 53 n. 2, 6i & n. i. 

Sea air, effects of, ll. 

Seals, mounting of, 91-92. 

Seasoning of wood, 117-18. 

Selenite, 301. 

Sellotape, 30, 167, 227. 

Sequins, 208. 

Seyrig, H., 226 & n. 

Shale, Kimmeridge, 324-j. 

Sheepskin, see parchment. 

Sheffield plate, 213. 

Shell, 156. 

Shellac, 142, 264, 304, 325, 330, 331. 

Shot-blasting, 204-5, 321-3. 

Sika deer, 155. 

Sihea gel; desiccation by, 3 1, 200, 245, 246, 
267, 336, 358- 

Siheaseal No. ia, 305 n. 

Siheates, in glass, 333-4. 

Sihceous materials, 198, 205, 207 246, 
294 ff. 

Silicones, 295. 

Siheon ester, spraying, 305. 

Silk, used in lamination, 60-61 ; and textde 
mounting, 106-7; m old embroidery, 
109; see also textiles. 

Silver, 188, 189, 190-1, 193, 195, 199, 212, 
232, 337; in gold alloy, 207, 208; gilding 
of, 209; decoration with ‘parcel gdt’, 
209; removal of tarnish from, 213-15; 
toughening of, 217-18, 223, 227, 228; 
alloys of, 220, 221, 224, 230; as decora- 
tion, 249, 250; test for, 251 & n.; solder- 
ing of, 288, 289. 

— Dip, 214, 231. 

— objects, 14, 212; cleaning of, 213-14; 
216-17, 230-1; exhibition of, 214-15; 
preservation ot patina on, 215-16; 
removal of dents from, 218-19; removal 
of copper corrosion from, 219 ff. ; electro- 
chemical reduction of, 222-3 ; electrolytic 

reduction of, 223-4; with iron incorpora- 
tions, 226-9; lacquering of, 229. 

Silver-plating, 289. 

Shnmons, R. H., 51 n. 2. 

Sizes, used for; canvas, 165 ; inks, 63 ; paper, 
51, 52, 54, 69, 83; papyrus, 42; resizing 
after repair, 87. 

Skin and skin products, 19 ff.; uses in 
ancient Egypt, 21; processing of, 21 ff.; 
fur skins, 23, 28, 29, 114; tawing of 24, 
25; oil dressing of 24; tanning of, 24-26, 
36-37; effect of moisture and moulds on, 
26-28, 31; insect attacks on, 28-30; pre- 
servation of 30-32, 34-39; repairing of 
32-33; unroUing of 39-40; see also 

Slate, 1 61. 

Snaking, 310. 

Soap, used with; bone and ivory, 147; 
leather, 33 n., 39; marble washing, 306; 
print cleaning, 73-74; textile cleaning, 97. 

Sodium chlorite, 59, 60, 77. 

Solderme, 288. 

Soldering, 286-9, 3 01. 

Solutions, preparation of 343, 349. 

Solvent Naphtha, 105. 

Solvents, used for cleaning; copper cor- 
rosion on silver, 219-22, 231; corroded 
metals, 186, 216, 257; cuprous enamel, 
338-9; lead and tin alloys, 270; pottery, 
327, 328, 330; prints, 78-81; rust, 277-9, 
291 ; shellac, 325 ; stained marble, 308-10; 
stone objects, 296-7, 300-1, 303; textiles, 
98 ff., 105, 108; varnishes, 72, 73, 174, 178. 

South Shields sword, 283. 

Spencer, E. W., 51 n. 2. 

Spraying, 123, 125, 133-4; of varnish, 177- 
8; of siheon ester on stone objects, 305-6. 

Stains; specific, removal of 78-84, 100-3, 
105, 219-22, 308-9, 312; as a guide to 
reconstruction, 146-7; on ancient monu- 
ments, 296; on marble, 307-10, 312-14. 

Stanwick scabbard, restoration of 138- 
40; sword, 282, No. 4. 

Statuary; treatment of 298, 303; weather- 
ing and, 316-17; packing of 317. 

Steel, 189, 195, 206, 271; stainless, 189 n., 

— objects, 206, 271, 274; treatment of 


277, 282-3, 290-1; protection of, 279; 
preservation of pattern in, 280-1. 

Stein, Sir Aurel, 50, 323. 

Sterilization; by fumigation, 55-57; against 
insects in wood, 123. 

Stoddard Solvent, 105. 

Stone objects, 295 ff.; cleaning and pre- 
servation of, 296 ff ; the paper pulp 
method and, 299-300; removal of in- 
soluble salts from, 300-1; consohdation 
of, 301-6; effects of weathering on, 3 16- 
17; repair of, 317-19- 

Stones, budding-, 12; deterioration of, 

Stoppings, 163, 167, 175, 176, 226, 

Storage of: copper alloys, 257; lead objects, 
264-5, 270; marble, 307-8; paper, 53- 
55; parchment, 49-50; silver, 214-15, 

Stout, G. L., 181. 

Strasbourg Museum, 13 n. 

Strip-lining, 166, 168. 

Stromberg, Elizabeth, no & n., 


Sulphuric acid; and leather, 36, 53, 1 12; and 
paper, 53, 112; in ink, 64; and canvas 
tendering, 166; and copper incrustation, 
239 ff.; and bronze, 243-4, 252; pre- 
paration of, 343; care of, 345. 

Sulphurous gases, 11-12, 19; and leather, 
34 ff., 53, 356; and paper, 52-53; and 
textiles, 112; and easel paintings, i66, 
1 70-1; and metals, 188. 

Sunhght, 9; and paper deterioration, 53; 
a bleaching agent, 75, 79; and textiles, 
94, 110-12; and waxed objects, 129; and 
bone and ivory, 145; and oil paintings, 


Surface enrichment, 207. 

Sutton Hoo finds, 26, 116, 209, 218, 251, 
282, 283; reconstruction of, 284-6. 

Swiss National Museum, 136 n. 2. 

Tablets, cuneiform; treatment of, 320-1; 
rendering them legible, 321-3; repair of, 


Tandberg, Prof. J., 215 n. 

Tannins, 24-25, 64. 


Tarnish: on silver, 213-15, 221, 222, 223; 
on copper, 232. 

Taylor, W. D., 67 & n. 3. 

Tears, repair of: in paper and parchment, 
86-88; in canvas, 167-8. 

Teepol detergent, 307, 327. 

Temperature changes; effects of, i ff., 55, 
129, 214; and textile deterioration, no; 
and painted panels, 160-1; and canvas 
paintings, 166; and stone objects, 302; 
reading of, 353. 

Tempering steel, 201 & n. 

Tendering of canvas, 166. 

Termites, 125-6. 

Terracotta, 327. 

Terylene, 104; and the mounting of firail 
textiles, 106-7, no. i33- 

Texibond V4N polyvinyl acetate emul- 
sion, 168 & n. I, 169. 

Textile Museum, Washington, D.C., 104, 

Textiles, 12, 14, 29, 93 ff.; for storage, 49- 
50; examination of, 94; cleaning of, 
95 ff., 98-100; washing of, 95-98; re- 
moval of stains from, 100-3, 105> 344! 
preservation of, 103-5; niould growths 
and, 93, 95, 106, 1 13; repair and mount- 
ing of, 106-10; hght and, 110-12; 
atmospheric pollution and, 1 12-13; 
insect attacks on, 113-15; isolation of, 
114; insecticides and, 115. 

Thacker, D. M. D., 279 n. 3. 

Thermometric equivalents, 352. 

Thymol, 56, 57-58, 73. 109. 113- 

Tiles, 329. 

Timber, see wood and wooden objects. 

Times, The, 12 n.; binding of, 34-35. 

Tin, 187, 188-9, 258, 265; in bronze, 233 
&n., 236, 247-8; as decoration, 249, 250, 
265; soldering of, 286-7. 

— objects, 262; cleaning of, 265; oxida- 
tion of, 266. 

— pest, 265-6. 

Todd, W., 151 n. & 2. 

Tools, used for: removing corrosion, 200- 
4; restoring gold objects, 21 1; repairing 
crushed silver, 218; restoring bronze 
objects, 254; cleaning marble, 314. 

Tortoise-shell, 155. 


Tower of London Armouries, 107 n. 

Transoptic resin for embedding objects, 

Tropics, the, 58, 70, 118, 126 n., 347. 

Tumbler tank, 225. 

Tune ship, 143. 

Turquoise, 294. 

Tut-ankhamiin’s tomb, 21, 208, 312. 

Upholstery, 19, 25; leather, preservation 
of, 33, 37; cleaning of, 98; sterilization 
of, 123. 

Ur of the Chaldees, 188-9, 211, 212, 216, 
217, 236. 

Vacuum treatment: against insects, 123; 
for consohdating wood, 130-1, 136, 137; 
for incnisted ivory, 150, 15 1; for in- 
crusted metals, 200, 223; for impregnat- 
ing stone objects, 304-5, 339. 

Vapour Phase Inhibitors, 279. 

Varnish, dammar and mastic, 172, 173, 


— removal of, 71-72, 173-5; for consohdat- 
ing wood, 129-30; for easel paintings, 
158, 170; deterioration of, 171, 172-3; 
reapplication of, 176 fF.; synthetic, 178- 
9; stabilization of, 359. 

Vellum, 44-45, 55 n. i; transparent, 46; 
cleaning of, 48-49; 83; sterilization of, 
57; Japanese, cleaning of, 82; removal of 
creases from, 85; repair of tears in, 87; 
mounting of, 88-89. 

Veneer, 132-3. 

Ventilation, 5, 8, 9, 27, 49, 353 ; and moulds, 
55, 93, 113, I2I. 

Vemon, W. H. J., 273 n. i. 

Versenes, for inhibiting rust, 278-9. 

Versenol, 312. 

Viking ships, 142-3. 

Vogt, E., 136 n. 2. 

Walbrook Mithraeum, 312. 

Walker, J. F., 55 & n. 2. 

Wallace Collection, 337. 

Wall-paintings, 323. 

Walls, dangers of waxing, 303. 

WaUs, H. J., 67 & n. 3. 

Warping: of wood, 118-19, i37> 140; 

of bone and ivory, 145, 146; in panel 
paintings, 160, 161; in canvas paintings, 
165; of marble, 310-11. 

Washing, methods used with: clay tablets, 
320-1; excavated pottery, 327-8; glass, 
335. 336, 338-9; marble, 306-7, 309; 
metalhc objects, 192, 196, 197-9, 222, 
231, 260, 270, 291; stones, 295, 298, 301. 

Waterlogging: and wood, 134 ff.; and 
bone and ivory, 145, 151. 

Waterproofing, 120, 143, 296-7, 303. 

Waters, C. E., 63 n. 

Wax, used for: removing bloom, 173 ; can- 
vas repairs, 169; impregnation, 127-30, 
133, 302-3, 311; iron and steel, 277, 279, 
280, 3 59 ; ivory" repairs, 150; lead preserva- 
tion, 259 ff . ; leather dressing for tropical 
use, 359 ; pewter, 269; pohshing, 206; re- 
vamishing, 179-80; wood sterilization, 
123, 125; preparation of microcrystalline 
polishing wax, 359. 

Weathering, 13; of stone, 294, 296; effect 
of, on stone objects, 316-17. 

Weeping glass, 336. 

Weiss, H. B., 58 n. i. 

Welwyn fire dogs, 282. 

Werner, A. E. A., 37 n., 176 n. i, 179. 

Westminster Abbey Effigies, 210. 

Westrosol dry-cleaning solvent, 99. 

Wheeler, Sir Mortimer, 138, 283 n. 2. 

Windsor Castle, Royal Collection at, 66. 

Wood, and paper-making, 51-52, 116; 
conditions under which it decays, 116- 
17; directional properties of, 117; season- 
ing of, 117-18; warping of, 118-19; 
fungus attacks on, 120-1; insect attack 
and, 121-3, 180; sterilization of, 123-5; 
proofing of, 125-6; strengthening of, 
126-31, 161-2; repair and maintenance 
of, 13 1-3; waterlogging and, 134-43: 
emitting tannic acid, 264-5. 

Wood, R. W., 208 & n. 

Wooden objects, 2, 8-9, 120; sterilization 
of, 124-5; proofing of, 126; consoHda- 
tion of, 127-9, 302; cleaning of, 133; 
waterlogged, preservation of, 134-43. 

Wood-worm, 121 ff., 130. 

Wool, ancient, cleaning of, 104-5; 



Woolley, Sir Leonard, 21 1. 

Workshop Notes, on the handling of ancient 
textiles, 104-5. 

Wright, E. V. and C. W., 141 n. 2. 

Writing: ground materials for, 41 ff.; pre- 
serving of, 53; faded, reading of, 66-67; 
clay tablets for, 319, 320; rendering it 
legible, 320-3; preservation of ostraka, 

X-ray examination: of ivory, 153; of 
paintings, 176; for hidden ornament, 
186; to penetrate incrustation, 220; of 
rust, 274, 275. 

York Minster, 121. 

Zinc, 189, 190; and electro-chemical reduc- 
tion, 191 ff. 




? r ‘**4 ii^k mcd is:shti'i» ^ 



Ay Depalrtment of Archaeology ’ '« 


o'. .%■ 

•. .•> 

Pleacfe he^ tis to. keep the boek: 
clean and moving. 

S. B« I4B.II.0EUH.